SlideShare a Scribd company logo

Multi-Robot Task Allocation using Meta-heuristic Optimization

Mohamed Ghanem
Mohamed Ghanem

The document discusses multi-robot task allocation using meta-heuristic optimization techniques, detailing the definition, applications, and challenges of multi-robot systems (MRS). It presents a proposed approach involving various algorithms such as tabu search and genetic algorithms, along with results and evaluation metrics. The conclusion outlines future work in improving task allocation efficiency and effectiveness.

Engineering
1 of 111
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
61
62
63
64
65
66
67
68
69
70
71
72
73
74
75
76
77
78
79
80
81
82
83
84
85
86
87
88
89
90
91
92
93
94
95
96
97
98
99
100
101
102
103
104
105
106
107
108
109
110
111
Introduction Multi-Robot Systems Task Allocation in MRS Proposed Approach Result & Discussion Conclusion & Future Work Multi-robot Task Allocation Using Meta-heuristic Optimization Mohamed Gomaa Ghanem Supervised by: Dr. Eng. Alaa Khamis German University in Cairo June 25, 2012 Mohamed Gomaa Ghanem German University in Cairo Multi-robot Task Allocation Using Meta-heuristic Optimization
Introduction Multi-Robot Systems Task Allocation in MRS Proposed Approach Result & Discussion Conclusion & Future Work Outline 1. Introduction Motivation Objective 2. Multi-robot System (MRS) Definition Application Challenging Aspects 3. Task Allocation in MRS Problem Formulation Architecture Approaches 4. Proposed Approach Meta-heuristic optimization techniques Tabu Search based Task Allocation Genetic Algorithm based Task Allocation Hybrid Approaches 5. Result & Discussion Experemental Setup Evaluation Metrics Results 6. Conclusion & Future Work Conclusion Future work Mohamed Gomaa Ghanem German University in Cairo Multi-robot Task Allocation Using Meta-heuristic Optimization
Introduction Multi-Robot Systems Task Allocation in MRS Proposed Approach Result & Discussion Conclusion & Future Work Outline 1. Introduction Motivation Objective 2. Multi-robot System (MRS) Definition Application Challenging Aspects 3. Task Allocation in MRS Problem Formulation Architecture Approaches 4. Proposed Approach Meta-heuristic optimization techniques Tabu Search based Task Allocation Genetic Algorithm based Task Allocation Hybrid Approaches 5. Result & Discussion Experemental Setup Evaluation Metrics Results 6. Conclusion & Future Work Conclusion Future work Mohamed Gomaa Ghanem German University in Cairo Multi-robot Task Allocation Using Meta-heuristic Optimization
Introduction Multi-Robot Systems Task Allocation in MRS Proposed Approach Result & Discussion Conclusion & Future Work Outline 1. Introduction Motivation Objective 2. Multi-robot System (MRS) Definition Application Challenging Aspects 3. Task Allocation in MRS Problem Formulation Architecture Approaches 4. Proposed Approach Meta-heuristic optimization techniques Tabu Search based Task Allocation Genetic Algorithm based Task Allocation Hybrid Approaches 5. Result & Discussion Experemental Setup Evaluation Metrics Results 6. Conclusion & Future Work Conclusion Future work Mohamed Gomaa Ghanem German University in Cairo Multi-robot Task Allocation Using Meta-heuristic Optimization
Introduction Multi-Robot Systems Task Allocation in MRS Proposed Approach Result & Discussion Conclusion & Future Work Outline 1. Introduction Motivation Objective 2. Multi-robot System (MRS) Definition Application Challenging Aspects 3. Task Allocation in MRS Problem Formulation Architecture Approaches 4. Proposed Approach Meta-heuristic optimization techniques Tabu Search based Task Allocation Genetic Algorithm based Task Allocation Hybrid Approaches 5. Result & Discussion Experemental Setup Evaluation Metrics Results 6. Conclusion & Future Work Conclusion Future work Mohamed Gomaa Ghanem German University in Cairo Multi-robot Task Allocation Using Meta-heuristic Optimization
Introduction Multi-Robot Systems Task Allocation in MRS Proposed Approach Result & Discussion Conclusion & Future Work Outline 1. Introduction Motivation Objective 2. Multi-robot System (MRS) Definition Application Challenging Aspects 3. Task Allocation in MRS Problem Formulation Architecture Approaches 4. Proposed Approach Meta-heuristic optimization techniques Tabu Search based Task Allocation Genetic Algorithm based Task Allocation Hybrid Approaches 5. Result & Discussion Experemental Setup Evaluation Metrics Results 6. Conclusion & Future Work Conclusion Future work Mohamed Gomaa Ghanem German University in Cairo Multi-robot Task Allocation Using Meta-heuristic Optimization
Introduction Multi-Robot Systems Task Allocation in MRS Proposed Approach Result & Discussion Conclusion & Future Work Outline 1. Introduction Motivation Objective 2. Multi-robot System (MRS) Definition Application Challenging Aspects 3. Task Allocation in MRS Problem Formulation Architecture Approaches 4. Proposed Approach Meta-heuristic optimization techniques Tabu Search based Task Allocation Genetic Algorithm based Task Allocation Hybrid Approaches 5. Result & Discussion Experemental Setup Evaluation Metrics Results 6. Conclusion & Future Work Conclusion Future work Mohamed Gomaa Ghanem German University in Cairo Multi-robot Task Allocation Using Meta-heuristic Optimization
Introduction Multi-Robot Systems Task Allocation in MRS Proposed Approach Result & Discussion Conclusion & Future Work Motivation 3/31 Who will do What by When? Mohamed Gomaa Ghanem German University in Cairo Multi-robot Task Allocation Using Meta-heuristic Optimization
Introduction Multi-Robot Systems Task Allocation in MRS Proposed Approach Result & Discussion Conclusion & Future Work Motivation 3/31 Who will do What by When? Mohamed Gomaa Ghanem German University in Cairo Multi-robot Task Allocation Using Meta-heuristic Optimization
Introduction Multi-Robot Systems Task Allocation in MRS Proposed Approach Result & Discussion Conclusion & Future Work Objective 4/31 Project Objectives Finding out which meta-heuristic is better in solving task allocation problem Implementing a framework that allows comparing and simulating a running algorithm(s) Mohamed Gomaa Ghanem German University in Cairo Multi-robot Task Allocation Using Meta-heuristic Optimization
Introduction Multi-Robot Systems Task Allocation in MRS Proposed Approach Result & Discussion Conclusion & Future Work Objective 4/31 Project Objectives Finding out which meta-heuristic is better in solving task allocation problem Implementing a framework that allows comparing and simulating a running algorithm(s) Mohamed Gomaa Ghanem German University in Cairo Multi-robot Task Allocation Using Meta-heuristic Optimization
Introduction Multi-Robot Systems Task Allocation in MRS Proposed Approach Result & Discussion Conclusion & Future Work Objective 4/31 Project Objectives Finding out which meta-heuristic is better in solving task allocation problem Implementing a framework that allows comparing and simulating a running algorithm(s) Mohamed Gomaa Ghanem German University in Cairo Multi-robot Task Allocation Using Meta-heuristic Optimization
Introduction Multi-Robot Systems Task Allocation in MRS Proposed Approach Result & Discussion Conclusion & Future Work Diffenetion 5/31 Multi-robot Systems are a group of robots that are designed aiming to perform some collection behavior Multi-robot System do resolve task complexity, increase performance, have more reliability, simple in design, and easily to be managed through a host computer. Mohamed Gomaa Ghanem German University in Cairo Multi-robot Task Allocation Using Meta-heuristic Optimization
Introduction Multi-Robot Systems Task Allocation in MRS Proposed Approach Result & Discussion Conclusion & Future Work Diffenetion 5/31 Multi-robot Systems are a group of robots that are designed aiming to perform some collection behavior Multi-robot System do resolve task complexity, increase performance, have more reliability, simple in design, and easily to be managed through a host computer. Mohamed Gomaa Ghanem German University in Cairo Multi-robot Task Allocation Using Meta-heuristic Optimization
Introduction Multi-Robot Systems Task Allocation in MRS Proposed Approach Result & Discussion Conclusion & Future Work Applications 6/31 The application of MRS includes, but is not limited to : Autonomous inspection of complex engineered structures Distributed sensing tasks in micro machinery or the human body Killing Cancer Tumors in Human Body Mining Agricultural Foraging Cooperative Tracking Surveillance, Reconnaissance and Intelligence Mohamed Gomaa Ghanem German University in Cairo Multi-robot Task Allocation Using Meta-heuristic Optimization
Introduction Multi-Robot Systems Task Allocation in MRS Proposed Approach Result & Discussion Conclusion & Future Work Challenging Aspects 7/31 MRS has many challenging aspects that represent an open research problems that could differ from one problem to another, Some of common challenging problem of MRS are: Analysis and Modeling the Problem Algorithm Design Implementation and Test Mohamed Gomaa Ghanem German University in Cairo Multi-robot Task Allocation Using Meta-heuristic Optimization
Introduction Multi-Robot Systems Task Allocation in MRS Proposed Approach Result & Discussion Conclusion & Future Work Challenging Aspects 7/31 MRS has many challenging aspects that represent an open research problems that could differ from one problem to another, Some of common challenging problem of MRS are: Analysis and Modeling the Problem Algorithm Design Implementation and Test Mohamed Gomaa Ghanem German University in Cairo Multi-robot Task Allocation Using Meta-heuristic Optimization
Introduction Multi-Robot Systems Task Allocation in MRS Proposed Approach Result & Discussion Conclusion & Future Work Challenging Aspects 7/31 MRS has many challenging aspects that represent an open research problems that could differ from one problem to another, Some of common challenging problem of MRS are: Analysis and Modeling the Problem Algorithm Design Implementation and Test Mohamed Gomaa Ghanem German University in Cairo Multi-robot Task Allocation Using Meta-heuristic Optimization
Introduction Multi-Robot Systems Task Allocation in MRS Proposed Approach Result & Discussion Conclusion & Future Work Challenging Aspects 7/31 MRS has many challenging aspects that represent an open research problems that could differ from one problem to another, Some of common challenging problem of MRS are: Analysis and Modeling the Problem Algorithm Design Implementation and Test Mohamed Gomaa Ghanem German University in Cairo Multi-robot Task Allocation Using Meta-heuristic Optimization
Introduction Multi-Robot Systems Task Allocation in MRS Proposed Approach Result & Discussion Conclusion & Future Work Challenging Aspects 7/31 MRS has many challenging aspects that represent an open research problems that could differ from one problem to another, Some of common challenging problem of MRS are: Analysis and Modeling the Problem Algorithm Design Implementation and Test Mohamed Gomaa Ghanem German University in Cairo Multi-robot Task Allocation Using Meta-heuristic Optimization
Introduction Multi-Robot Systems Task Allocation in MRS Proposed Approach Result & Discussion Conclusion & Future Work Problem Formulation 8/31 Task allocation is a NP hard problem that addresses how to optimally assign a set of tasks to a set of robots to maximize overall expected performance, taking into account the priorities of the tasks and the skill ratings of the robots. The goal is to find out which algorithm is better in allocating N number of robots to M number of tasks laying on an area of A2 and reaching the minimum Travel Distance and/or Mission Completion Time for all robots to cover all required tasks. Mohamed Gomaa Ghanem German University in Cairo Multi-robot Task Allocation Using Meta-heuristic Optimization
Introduction Multi-Robot Systems Task Allocation in MRS Proposed Approach Result & Discussion Conclusion & Future Work Problem Formulation 8/31 Task allocation is a NP hard problem that addresses how to optimally assign a set of tasks to a set of robots to maximize overall expected performance, taking into account the priorities of the tasks and the skill ratings of the robots. The goal is to find out which algorithm is better in allocating N number of robots to M number of tasks laying on an area of A2 and reaching the minimum Travel Distance and/or Mission Completion Time for all robots to cover all required tasks. Mohamed Gomaa Ghanem German University in Cairo Multi-robot Task Allocation Using Meta-heuristic Optimization
Introduction Multi-Robot Systems Task Allocation in MRS Proposed Approach Result & Discussion Conclusion & Future Work Architecture Centralized Architecture Robot team treated as a single ”system” with many degrees of freedom. A single robot is the ”leader”, which plans optimal actions for group. Group members send information to leader and carry out actions. Pros Cons Leader can take all relevant information into ac- count. Computationally hard and sometimes intractable for more than a few robots. In theory, coordination can be perfect: Optimal plans possible. Makes unrealistic assumptions, where all relevant info can be transmitted to leader, and this info doesn’t change during plan construction. Result: response sluggish or inaccurate Vulnerable to malfunction of leader Heavy communication load Mohamed Gomaa Ghanem German University in Cairo Multi-robot Task Allocation Using Meta-heuristic Optimization
Introduction Multi-Robot Systems Task Allocation in MRS Proposed Approach Result & Discussion Conclusion & Future Work Architecture Centralized Architecture Robot team treated as a single ”system” with many degrees of freedom. A single robot is the ”leader”, which plans optimal actions for group. Group members send information to leader and carry out actions. Pros Cons Leader can take all relevant information into ac- count. Computationally hard and sometimes intractable for more than a few robots. In theory, coordination can be perfect: Optimal plans possible. Makes unrealistic assumptions, where all relevant info can be transmitted to leader, and this info doesn’t change during plan construction. Result: response sluggish or inaccurate Vulnerable to malfunction of leader Heavy communication load Mohamed Gomaa Ghanem German University in Cairo Multi-robot Task Allocation Using Meta-heuristic Optimization
Introduction Multi-Robot Systems Task Allocation in MRS Proposed Approach Result & Discussion Conclusion & Future Work Architecture Decentralized Architecture Distributed architecture is concentrated in planning responsibility spread over team, where each robot basically independent from the others and robots use locally observable information to make their plans. Pros Cons Fast response to dynamic conditions. Not all problems can be decomposed well. Little or no communication required. Plans are based only on local information. Little computation required. Result: solutions are often highly suboptimal. Smooth response to environmental changes. Mohamed Gomaa Ghanem German University in Cairo Multi-robot Task Allocation Using Meta-heuristic Optimization
Introduction Multi-Robot Systems Task Allocation in MRS Proposed Approach Result & Discussion Conclusion & Future Work Architecture Decentralized Architecture Distributed architecture is concentrated in planning responsibility spread over team, where each robot basically independent from the others and robots use locally observable information to make their plans. Pros Cons Fast response to dynamic conditions. Not all problems can be decomposed well. Little or no communication required. Plans are based only on local information. Little computation required. Result: solutions are often highly suboptimal. Smooth response to environmental changes. Mohamed Gomaa Ghanem German University in Cairo Multi-robot Task Allocation Using Meta-heuristic Optimization
Introduction Multi-Robot Systems Task Allocation in MRS Proposed Approach Result & Discussion Conclusion & Future Work Approaches 11/31 Market Based Approaches Market-based approach is based on the economic model of a free market, each robot seeks to maximize individual ”profit”, Robots can negotiate and bid for tasks individual profit helps the common good, and decisions are made locally but effects approach optimality. Pros Cons Robustness and quickness of distributed sys- tem. Cost heuristics can be inaccurate. Approaches optimality of centralized sys- tem. Much of this approach is still under development. Low communication requirements. Mohamed Gomaa Ghanem German University in Cairo Multi-robot Task Allocation Using Meta-heuristic Optimization
Introduction Multi-Robot Systems Task Allocation in MRS Proposed Approach Result & Discussion Conclusion & Future Work Approaches 11/31 Market Based Approaches Market-based approach is based on the economic model of a free market, each robot seeks to maximize individual ”profit”, Robots can negotiate and bid for tasks individual profit helps the common good, and decisions are made locally but effects approach optimality. Pros Cons Robustness and quickness of distributed sys- tem. Cost heuristics can be inaccurate. Approaches optimality of centralized sys- tem. Much of this approach is still under development. Low communication requirements. Mohamed Gomaa Ghanem German University in Cairo Multi-robot Task Allocation Using Meta-heuristic Optimization
Introduction Multi-Robot Systems Task Allocation in MRS Proposed Approach Result & Discussion Conclusion & Future Work Approaches 12/31 Optimization Based Approaches Response Surface Methodology. Gradient-Based Search. Heuristic Searches. Mohamed Gomaa Ghanem German University in Cairo Multi-robot Task Allocation Using Meta-heuristic Optimization
Introduction Multi-Robot Systems Task Allocation in MRS Proposed Approach Result & Discussion Conclusion & Future Work Approaches 12/31 Optimization Based Approaches Response Surface Methodology. Gradient-Based Search. Heuristic Searches. Mohamed Gomaa Ghanem German University in Cairo Multi-robot Task Allocation Using Meta-heuristic Optimization
Introduction Multi-Robot Systems Task Allocation in MRS Proposed Approach Result & Discussion Conclusion & Future Work Approaches 12/31 Optimization Based Approaches Response Surface Methodology. Gradient-Based Search. Heuristic Searches. Mohamed Gomaa Ghanem German University in Cairo Multi-robot Task Allocation Using Meta-heuristic Optimization
Introduction Multi-Robot Systems Task Allocation in MRS Proposed Approach Result & Discussion Conclusion & Future Work Approaches 12/31 Optimization Based Approaches Response Surface Methodology. Gradient-Based Search. Heuristic Searches. Mohamed Gomaa Ghanem German University in Cairo Multi-robot Task Allocation Using Meta-heuristic Optimization
Introduction Multi-Robot Systems Task Allocation in MRS Proposed Approach Result & Discussion Conclusion & Future Work Meta-heuristic optimization techniques 13/31 Meta-heuristic classifications Main classification Used Algorithms Mohamed Gomaa Ghanem German University in Cairo Multi-robot Task Allocation Using Meta-heuristic Optimization
Introduction Multi-Robot Systems Task Allocation in MRS Proposed Approach Result & Discussion Conclusion & Future Work Meta-heuristic optimization techniques 13/31 Meta-heuristic classifications Main classification Used Algorithms Mohamed Gomaa Ghanem German University in Cairo Multi-robot Task Allocation Using Meta-heuristic Optimization
Introduction Multi-Robot Systems Task Allocation in MRS Proposed Approach Result & Discussion Conclusion & Future Work Meta-heuristic optimization techniques 13/31 Meta-heuristic classifications Main classification Used Algorithms Mohamed Gomaa Ghanem German University in Cairo Multi-robot Task Allocation Using Meta-heuristic Optimization
Introduction Multi-Robot Systems Task Allocation in MRS Proposed Approach Result & Discussion Conclusion & Future Work Tabu Search based Task Allocation 14/31 TS Properties Trajectory based Local search With memory Naturally inspired Pros Cons Allow non-improving solution to accept in order to escape from local optimum. Global optimum may not be found, depends on parameter settings. Can be applied to both discrete and contin- uous solution spaces. Can obtain solutions that often surpass a best solution previously found by other ap- proaches. Mohamed Gomaa Ghanem German University in Cairo Multi-robot Task Allocation Using Meta-heuristic Optimization
Introduction Multi-Robot Systems Task Allocation in MRS Proposed Approach Result & Discussion Conclusion & Future Work Tabu Search based Task Allocation 14/31 TS Properties Trajectory based Local search With memory Naturally inspired Pros Cons Allow non-improving solution to accept in order to escape from local optimum. Global optimum may not be found, depends on parameter settings. Can be applied to both discrete and contin- uous solution spaces. Can obtain solutions that often surpass a best solution previously found by other ap- proaches. Mohamed Gomaa Ghanem German University in Cairo Multi-robot Task Allocation Using Meta-heuristic Optimization
Introduction Multi-Robot Systems Task Allocation in MRS Proposed Approach Result & Discussion Conclusion & Future Work Tabu Search based Task Allocation 14/31 TS Properties Trajectory based Local search With memory Naturally inspired Pros Cons Allow non-improving solution to accept in order to escape from local optimum. Global optimum may not be found, depends on parameter settings. Can be applied to both discrete and contin- uous solution spaces. Can obtain solutions that often surpass a best solution previously found by other ap- proaches. Mohamed Gomaa Ghanem German University in Cairo Multi-robot Task Allocation Using Meta-heuristic Optimization
Introduction Multi-Robot Systems Task Allocation in MRS Proposed Approach Result & Discussion Conclusion & Future Work Tabu Search based Task Allocation 14/31 TS Properties Trajectory based Local search With memory Naturally inspired Pros Cons Allow non-improving solution to accept in order to escape from local optimum. Global optimum may not be found, depends on parameter settings. Can be applied to both discrete and contin- uous solution spaces. Can obtain solutions that often surpass a best solution previously found by other ap- proaches. Mohamed Gomaa Ghanem German University in Cairo Multi-robot Task Allocation Using Meta-heuristic Optimization
Introduction Multi-Robot Systems Task Allocation in MRS Proposed Approach Result & Discussion Conclusion & Future Work Tabu Search based Task Allocation 14/31 TS Properties Trajectory based Local search With memory Naturally inspired Pros Cons Allow non-improving solution to accept in order to escape from local optimum. Global optimum may not be found, depends on parameter settings. Can be applied to both discrete and contin- uous solution spaces. Can obtain solutions that often surpass a best solution previously found by other ap- proaches. Mohamed Gomaa Ghanem German University in Cairo Multi-robot Task Allocation Using Meta-heuristic Optimization
Introduction Multi-Robot Systems Task Allocation in MRS Proposed Approach Result & Discussion Conclusion & Future Work Tabu Search based Task Allocation 14/31 TS Properties Trajectory based Local search With memory Naturally inspired Pros Cons Allow non-improving solution to accept in order to escape from local optimum. Global optimum may not be found, depends on parameter settings. Can be applied to both discrete and contin- uous solution spaces. Can obtain solutions that often surpass a best solution previously found by other ap- proaches. Mohamed Gomaa Ghanem German University in Cairo Multi-robot Task Allocation Using Meta-heuristic Optimization
Introduction Multi-Robot Systems Task Allocation in MRS Proposed Approach Result & Discussion Conclusion & Future Work Tabu Search based Task Allocation 15/31 TS Mechanism Mohamed Gomaa Ghanem German University in Cairo Multi-robot Task Allocation Using Meta-heuristic Optimization
Introduction Multi-Robot Systems Task Allocation in MRS Proposed Approach Result & Discussion Conclusion & Future Work Genetic Algorithm based Task Allocation 16/31 GA Properties Population based With memory Naturally inspired Pros Cons Often locate good solutions. Time Delay. This is an effective heuristic when dealing with a very large solution space. Tend to converge towards local points, rather than global points Mutation introduces new information gene pool, that protects against converging too quickly to local optimum. Mohamed Gomaa Ghanem German University in Cairo Multi-robot Task Allocation Using Meta-heuristic Optimization
Introduction Multi-Robot Systems Task Allocation in MRS Proposed Approach Result & Discussion Conclusion & Future Work Genetic Algorithm based Task Allocation 16/31 GA Properties Population based With memory Naturally inspired Pros Cons Often locate good solutions. Time Delay. This is an effective heuristic when dealing with a very large solution space. Tend to converge towards local points, rather than global points Mutation introduces new information gene pool, that protects against converging too quickly to local optimum. Mohamed Gomaa Ghanem German University in Cairo Multi-robot Task Allocation Using Meta-heuristic Optimization
Introduction Multi-Robot Systems Task Allocation in MRS Proposed Approach Result & Discussion Conclusion & Future Work Genetic Algorithm based Task Allocation 16/31 GA Properties Population based With memory Naturally inspired Pros Cons Often locate good solutions. Time Delay. This is an effective heuristic when dealing with a very large solution space. Tend to converge towards local points, rather than global points Mutation introduces new information gene pool, that protects against converging too quickly to local optimum. Mohamed Gomaa Ghanem German University in Cairo Multi-robot Task Allocation Using Meta-heuristic Optimization
Introduction Multi-Robot Systems Task Allocation in MRS Proposed Approach Result & Discussion Conclusion & Future Work Genetic Algorithm based Task Allocation 16/31 GA Properties Population based With memory Naturally inspired Pros Cons Often locate good solutions. Time Delay. This is an effective heuristic when dealing with a very large solution space. Tend to converge towards local points, rather than global points Mutation introduces new information gene pool, that protects against converging too quickly to local optimum. Mohamed Gomaa Ghanem German University in Cairo Multi-robot Task Allocation Using Meta-heuristic Optimization
Introduction Multi-Robot Systems Task Allocation in MRS Proposed Approach Result & Discussion Conclusion & Future Work Genetic Algorithm based Task Allocation 16/31 GA Properties Population based With memory Naturally inspired Pros Cons Often locate good solutions. Time Delay. This is an effective heuristic when dealing with a very large solution space. Tend to converge towards local points, rather than global points Mutation introduces new information gene pool, that protects against converging too quickly to local optimum. Mohamed Gomaa Ghanem German University in Cairo Multi-robot Task Allocation Using Meta-heuristic Optimization
Introduction Multi-Robot Systems Task Allocation in MRS Proposed Approach Result & Discussion Conclusion & Future Work Genetic Algorithm based Task Allocation 17/31 GA Mechanism Mohamed Gomaa Ghanem German University in Cairo Multi-robot Task Allocation Using Meta-heuristic Optimization
Introduction Multi-Robot Systems Task Allocation in MRS Proposed Approach Result & Discussion Conclusion & Future Work Hybrid Approaches 18/31 Two Hybrid Approaches TS-GA:: best solution from TS to be considered as an intial solution in GA Require: IntialSolution & TasksPosition & RobotsPosition Solutionbest ← Tabu-Search(IntialSolution,TasksPosition,RobotsPosition) RUN Genetic-Algorithm(Solutionbest ,TasksPosition,RobotsPosition) return Best achieved solution GA-TS:: best solution from GA to be considered as an intial solution in TS Require: IntialSolution & TasksPosition & RobotsPosition Solutionbest ← Genetic-Algorithm(IntialSolution,TasksPosition,RobotsPosition) RUN Tabu-Search(Solutionbest ,TasksPosition,RobotsPosition) return Best achieved solution Mohamed Gomaa Ghanem German University in Cairo Multi-robot Task Allocation Using Meta-heuristic Optimization
Introduction Multi-Robot Systems Task Allocation in MRS Proposed Approach Result & Discussion Conclusion & Future Work Hybrid Approaches 18/31 Two Hybrid Approaches TS-GA:: best solution from TS to be considered as an intial solution in GA Require: IntialSolution & TasksPosition & RobotsPosition Solutionbest ← Tabu-Search(IntialSolution,TasksPosition,RobotsPosition) RUN Genetic-Algorithm(Solutionbest ,TasksPosition,RobotsPosition) return Best achieved solution GA-TS:: best solution from GA to be considered as an intial solution in TS Require: IntialSolution & TasksPosition & RobotsPosition Solutionbest ← Genetic-Algorithm(IntialSolution,TasksPosition,RobotsPosition) RUN Tabu-Search(Solutionbest ,TasksPosition,RobotsPosition) return Best achieved solution Mohamed Gomaa Ghanem German University in Cairo Multi-robot Task Allocation Using Meta-heuristic Optimization
Introduction Multi-Robot Systems Task Allocation in MRS Proposed Approach Result & Discussion Conclusion & Future Work Hybrid Approaches 18/31 Two Hybrid Approaches TS-GA:: best solution from TS to be considered as an intial solution in GA Require: IntialSolution & TasksPosition & RobotsPosition Solutionbest ← Tabu-Search(IntialSolution,TasksPosition,RobotsPosition) RUN Genetic-Algorithm(Solutionbest ,TasksPosition,RobotsPosition) return Best achieved solution GA-TS:: best solution from GA to be considered as an intial solution in TS Require: IntialSolution & TasksPosition & RobotsPosition Solutionbest ← Genetic-Algorithm(IntialSolution,TasksPosition,RobotsPosition) RUN Tabu-Search(Solutionbest ,TasksPosition,RobotsPosition) return Best achieved solution Mohamed Gomaa Ghanem German University in Cairo Multi-robot Task Allocation Using Meta-heuristic Optimization
Introduction Multi-Robot Systems Task Allocation in MRS Proposed Approach Result & Discussion Conclusion & Future Work Hybrid Approaches 18/31 Two Hybrid Approaches TS-GA:: best solution from TS to be considered as an intial solution in GA Require: IntialSolution & TasksPosition & RobotsPosition Solutionbest ← Tabu-Search(IntialSolution,TasksPosition,RobotsPosition) RUN Genetic-Algorithm(Solutionbest ,TasksPosition,RobotsPosition) return Best achieved solution GA-TS:: best solution from GA to be considered as an intial solution in TS Require: IntialSolution & TasksPosition & RobotsPosition Solutionbest ← Genetic-Algorithm(IntialSolution,TasksPosition,RobotsPosition) RUN Tabu-Search(Solutionbest ,TasksPosition,RobotsPosition) return Best achieved solution Mohamed Gomaa Ghanem German University in Cairo Multi-robot Task Allocation Using Meta-heuristic Optimization
Introduction Multi-Robot Systems Task Allocation in MRS Proposed Approach Result & Discussion Conclusion & Future Work Hybrid Approaches 18/31 Two Hybrid Approaches TS-GA:: best solution from TS to be considered as an intial solution in GA Require: IntialSolution & TasksPosition & RobotsPosition Solutionbest ← Tabu-Search(IntialSolution,TasksPosition,RobotsPosition) RUN Genetic-Algorithm(Solutionbest ,TasksPosition,RobotsPosition) return Best achieved solution GA-TS:: best solution from GA to be considered as an intial solution in TS Require: IntialSolution & TasksPosition & RobotsPosition Solutionbest ← Genetic-Algorithm(IntialSolution,TasksPosition,RobotsPosition) RUN Tabu-Search(Solutionbest ,TasksPosition,RobotsPosition) return Best achieved solution Mohamed Gomaa Ghanem German University in Cairo Multi-robot Task Allocation Using Meta-heuristic Optimization
Introduction Multi-Robot Systems Task Allocation in MRS Proposed Approach Result & Discussion Conclusion & Future Work Hybrid Approaches 18/31 Two Hybrid Approaches TS-GA:: best solution from TS to be considered as an intial solution in GA Require: IntialSolution & TasksPosition & RobotsPosition Solutionbest ← Tabu-Search(IntialSolution,TasksPosition,RobotsPosition) RUN Genetic-Algorithm(Solutionbest ,TasksPosition,RobotsPosition) return Best achieved solution GA-TS:: best solution from GA to be considered as an intial solution in TS Require: IntialSolution & TasksPosition & RobotsPosition Solutionbest ← Genetic-Algorithm(IntialSolution,TasksPosition,RobotsPosition) RUN Tabu-Search(Solutionbest ,TasksPosition,RobotsPosition) return Best achieved solution Mohamed Gomaa Ghanem German University in Cairo Multi-robot Task Allocation Using Meta-heuristic Optimization
Introduction Multi-Robot Systems Task Allocation in MRS Proposed Approach Result & Discussion Conclusion & Future Work Hybrid Approaches 18/31 Two Hybrid Approaches TS-GA:: best solution from TS to be considered as an intial solution in GA Require: IntialSolution & TasksPosition & RobotsPosition Solutionbest ← Tabu-Search(IntialSolution,TasksPosition,RobotsPosition) RUN Genetic-Algorithm(Solutionbest ,TasksPosition,RobotsPosition) return Best achieved solution GA-TS:: best solution from GA to be considered as an intial solution in TS Require: IntialSolution & TasksPosition & RobotsPosition Solutionbest ← Genetic-Algorithm(IntialSolution,TasksPosition,RobotsPosition) RUN Tabu-Search(Solutionbest ,TasksPosition,RobotsPosition) return Best achieved solution Mohamed Gomaa Ghanem German University in Cairo Multi-robot Task Allocation Using Meta-heuristic Optimization
Introduction Multi-Robot Systems Task Allocation in MRS Proposed Approach Result & Discussion Conclusion & Future Work Hybrid Approaches 18/31 Two Hybrid Approaches TS-GA:: best solution from TS to be considered as an intial solution in GA Require: IntialSolution & TasksPosition & RobotsPosition Solutionbest ← Tabu-Search(IntialSolution,TasksPosition,RobotsPosition) RUN Genetic-Algorithm(Solutionbest ,TasksPosition,RobotsPosition) return Best achieved solution GA-TS:: best solution from GA to be considered as an intial solution in TS Require: IntialSolution & TasksPosition & RobotsPosition Solutionbest ← Genetic-Algorithm(IntialSolution,TasksPosition,RobotsPosition) RUN Tabu-Search(Solutionbest ,TasksPosition,RobotsPosition) return Best achieved solution Mohamed Gomaa Ghanem German University in Cairo Multi-robot Task Allocation Using Meta-heuristic Optimization
Introduction Multi-Robot Systems Task Allocation in MRS Proposed Approach Result & Discussion Conclusion & Future Work Hybrid Approaches 18/31 Two Hybrid Approaches TS-GA:: best solution from TS to be considered as an intial solution in GA Require: IntialSolution & TasksPosition & RobotsPosition Solutionbest ← Tabu-Search(IntialSolution,TasksPosition,RobotsPosition) RUN Genetic-Algorithm(Solutionbest ,TasksPosition,RobotsPosition) return Best achieved solution GA-TS:: best solution from GA to be considered as an intial solution in TS Require: IntialSolution & TasksPosition & RobotsPosition Solutionbest ← Genetic-Algorithm(IntialSolution,TasksPosition,RobotsPosition) RUN Tabu-Search(Solutionbest ,TasksPosition,RobotsPosition) return Best achieved solution Mohamed Gomaa Ghanem German University in Cairo Multi-robot Task Allocation Using Meta-heuristic Optimization
Introduction Multi-Robot Systems Task Allocation in MRS Proposed Approach Result & Discussion Conclusion & Future Work Hybrid Approaches 18/31 Two Hybrid Approaches TS-GA:: best solution from TS to be considered as an intial solution in GA Require: IntialSolution & TasksPosition & RobotsPosition Solutionbest ← Tabu-Search(IntialSolution,TasksPosition,RobotsPosition) RUN Genetic-Algorithm(Solutionbest ,TasksPosition,RobotsPosition) return Best achieved solution GA-TS:: best solution from GA to be considered as an intial solution in TS Require: IntialSolution & TasksPosition & RobotsPosition Solutionbest ← Genetic-Algorithm(IntialSolution,TasksPosition,RobotsPosition) RUN Tabu-Search(Solutionbest ,TasksPosition,RobotsPosition) return Best achieved solution Mohamed Gomaa Ghanem German University in Cairo Multi-robot Task Allocation Using Meta-heuristic Optimization
Introduction Multi-Robot Systems Task Allocation in MRS Proposed Approach Result & Discussion Conclusion & Future Work Hybrid Approaches 18/31 Two Hybrid Approaches TS-GA:: best solution from TS to be considered as an intial solution in GA Require: IntialSolution & TasksPosition & RobotsPosition Solutionbest ← Tabu-Search(IntialSolution,TasksPosition,RobotsPosition) RUN Genetic-Algorithm(Solutionbest ,TasksPosition,RobotsPosition) return Best achieved solution GA-TS:: best solution from GA to be considered as an intial solution in TS Require: IntialSolution & TasksPosition & RobotsPosition Solutionbest ← Genetic-Algorithm(IntialSolution,TasksPosition,RobotsPosition) RUN Tabu-Search(Solutionbest ,TasksPosition,RobotsPosition) return Best achieved solution Mohamed Gomaa Ghanem German University in Cairo Multi-robot Task Allocation Using Meta-heuristic Optimization
Introduction Multi-Robot Systems Task Allocation in MRS Proposed Approach Result & Discussion Conclusion & Future Work Hybrid Approaches 18/31 Two Hybrid Approaches TS-GA:: best solution from TS to be considered as an intial solution in GA Require: IntialSolution & TasksPosition & RobotsPosition Solutionbest ← Tabu-Search(IntialSolution,TasksPosition,RobotsPosition) RUN Genetic-Algorithm(Solutionbest ,TasksPosition,RobotsPosition) return Best achieved solution GA-TS:: best solution from GA to be considered as an intial solution in TS Require: IntialSolution & TasksPosition & RobotsPosition Solutionbest ← Genetic-Algorithm(IntialSolution,TasksPosition,RobotsPosition) RUN Tabu-Search(Solutionbest ,TasksPosition,RobotsPosition) return Best achieved solution Mohamed Gomaa Ghanem German University in Cairo Multi-robot Task Allocation Using Meta-heuristic Optimization
Introduction Multi-Robot Systems Task Allocation in MRS Proposed Approach Result & Discussion Conclusion & Future Work Hybrid Approaches 18/31 Two Hybrid Approaches TS-GA:: best solution from TS to be considered as an intial solution in GA Require: IntialSolution & TasksPosition & RobotsPosition Solutionbest ← Tabu-Search(IntialSolution,TasksPosition,RobotsPosition) RUN Genetic-Algorithm(Solutionbest ,TasksPosition,RobotsPosition) return Best achieved solution GA-TS:: best solution from GA to be considered as an intial solution in TS Require: IntialSolution & TasksPosition & RobotsPosition Solutionbest ← Genetic-Algorithm(IntialSolution,TasksPosition,RobotsPosition) RUN Tabu-Search(Solutionbest ,TasksPosition,RobotsPosition) return Best achieved solution Mohamed Gomaa Ghanem German University in Cairo Multi-robot Task Allocation Using Meta-heuristic Optimization
Introduction Multi-Robot Systems Task Allocation in MRS Proposed Approach Result & Discussion Conclusion & Future Work Experemental Setup 19/31 In all of experiments,some component were used: Java4MRS Simbad-3D Simulator OpenTS & Jgap Matlab The result is an output of running the above programs on a machine with: 32-bit windows Operating System AMD Turion 64 X2 Mobile Technology TL-68 2.4 GHz Processor 3.00 GB of RAM Mohamed Gomaa Ghanem German University in Cairo Multi-robot Task Allocation Using Meta-heuristic Optimization
Introduction Multi-Robot Systems Task Allocation in MRS Proposed Approach Result & Discussion Conclusion & Future Work Experemental Setup 19/31 In all of experiments,some component were used: Java4MRS Simbad-3D Simulator OpenTS & Jgap Matlab The result is an output of running the above programs on a machine with: 32-bit windows Operating System AMD Turion 64 X2 Mobile Technology TL-68 2.4 GHz Processor 3.00 GB of RAM Mohamed Gomaa Ghanem German University in Cairo Multi-robot Task Allocation Using Meta-heuristic Optimization
Introduction Multi-Robot Systems Task Allocation in MRS Proposed Approach Result & Discussion Conclusion & Future Work Experemental Setup 19/31 In all of experiments,some component were used: Java4MRS Simbad-3D Simulator OpenTS & Jgap Matlab The result is an output of running the above programs on a machine with: 32-bit windows Operating System AMD Turion 64 X2 Mobile Technology TL-68 2.4 GHz Processor 3.00 GB of RAM Mohamed Gomaa Ghanem German University in Cairo Multi-robot Task Allocation Using Meta-heuristic Optimization
Introduction Multi-Robot Systems Task Allocation in MRS Proposed Approach Result & Discussion Conclusion & Future Work Experemental Setup 19/31 In all of experiments,some component were used: Java4MRS Simbad-3D Simulator OpenTS & Jgap Matlab The result is an output of running the above programs on a machine with: 32-bit windows Operating System AMD Turion 64 X2 Mobile Technology TL-68 2.4 GHz Processor 3.00 GB of RAM Mohamed Gomaa Ghanem German University in Cairo Multi-robot Task Allocation Using Meta-heuristic Optimization
Introduction Multi-Robot Systems Task Allocation in MRS Proposed Approach Result & Discussion Conclusion & Future Work Experemental Setup 19/31 In all of experiments,some component were used: Java4MRS Simbad-3D Simulator OpenTS & Jgap Matlab The result is an output of running the above programs on a machine with: 32-bit windows Operating System AMD Turion 64 X2 Mobile Technology TL-68 2.4 GHz Processor 3.00 GB of RAM Mohamed Gomaa Ghanem German University in Cairo Multi-robot Task Allocation Using Meta-heuristic Optimization
Introduction Multi-Robot Systems Task Allocation in MRS Proposed Approach Result & Discussion Conclusion & Future Work Experemental Setup 19/31 In all of experiments,some component were used: Java4MRS Simbad-3D Simulator OpenTS & Jgap Matlab The result is an output of running the above programs on a machine with: 32-bit windows Operating System AMD Turion 64 X2 Mobile Technology TL-68 2.4 GHz Processor 3.00 GB of RAM Mohamed Gomaa Ghanem German University in Cairo Multi-robot Task Allocation Using Meta-heuristic Optimization
Introduction Multi-Robot Systems Task Allocation in MRS Proposed Approach Result & Discussion Conclusion & Future Work Experemental Setup 19/31 In all of experiments,some component were used: Java4MRS Simbad-3D Simulator OpenTS & Jgap Matlab The result is an output of running the above programs on a machine with: 32-bit windows Operating System AMD Turion 64 X2 Mobile Technology TL-68 2.4 GHz Processor 3.00 GB of RAM Mohamed Gomaa Ghanem German University in Cairo Multi-robot Task Allocation Using Meta-heuristic Optimization
Introduction Multi-Robot Systems Task Allocation in MRS Proposed Approach Result & Discussion Conclusion & Future Work Experemental Setup 19/31 In all of experiments,some component were used: Java4MRS Simbad-3D Simulator OpenTS & Jgap Matlab The result is an output of running the above programs on a machine with: 32-bit windows Operating System AMD Turion 64 X2 Mobile Technology TL-68 2.4 GHz Processor 3.00 GB of RAM Mohamed Gomaa Ghanem German University in Cairo Multi-robot Task Allocation Using Meta-heuristic Optimization
Introduction Multi-Robot Systems Task Allocation in MRS Proposed Approach Result & Discussion Conclusion & Future Work Experemental Setup 19/31 In all of experiments,some component were used: Java4MRS Simbad-3D Simulator OpenTS & Jgap Matlab The result is an output of running the above programs on a machine with: 32-bit windows Operating System AMD Turion 64 X2 Mobile Technology TL-68 2.4 GHz Processor 3.00 GB of RAM Mohamed Gomaa Ghanem German University in Cairo Multi-robot Task Allocation Using Meta-heuristic Optimization
Introduction Multi-Robot Systems Task Allocation in MRS Proposed Approach Result & Discussion Conclusion & Future Work Evaluation Metrics 20/31 Cost of traveled distance Cost = numberOfRobots i=0 [dist(Ri , T0) + numOfTasks−1forRi j=0 dist(Tj, Tj+1)] Mission completion time = Algorithm Running time + Time taken by last robot to finish its tasks Reliability Scalability Each will be evaluated according to some variables: Number of robots Number of tasks Map area size Algorithm’s parameters Mohamed Gomaa Ghanem German University in Cairo Multi-robot Task Allocation Using Meta-heuristic Optimization
Introduction Multi-Robot Systems Task Allocation in MRS Proposed Approach Result & Discussion Conclusion & Future Work Evaluation Metrics 20/31 Cost of traveled distance Cost = numberOfRobots i=0 [dist(Ri , T0) + numOfTasks−1forRi j=0 dist(Tj, Tj+1)] Mission completion time = Algorithm Running time + Time taken by last robot to finish its tasks Reliability Scalability Each will be evaluated according to some variables: Number of robots Number of tasks Map area size Algorithm’s parameters Mohamed Gomaa Ghanem German University in Cairo Multi-robot Task Allocation Using Meta-heuristic Optimization
Introduction Multi-Robot Systems Task Allocation in MRS Proposed Approach Result & Discussion Conclusion & Future Work Evaluation Metrics 20/31 Cost of traveled distance Cost = numberOfRobots i=0 [dist(Ri , T0) + numOfTasks−1forRi j=0 dist(Tj, Tj+1)] Mission completion time = Algorithm Running time + Time taken by last robot to finish its tasks Reliability Scalability Each will be evaluated according to some variables: Number of robots Number of tasks Map area size Algorithm’s parameters Mohamed Gomaa Ghanem German University in Cairo Multi-robot Task Allocation Using Meta-heuristic Optimization
Introduction Multi-Robot Systems Task Allocation in MRS Proposed Approach Result & Discussion Conclusion & Future Work Evaluation Metrics 20/31 Cost of traveled distance Cost = numberOfRobots i=0 [dist(Ri , T0) + numOfTasks−1forRi j=0 dist(Tj, Tj+1)] Mission completion time = Algorithm Running time + Time taken by last robot to finish its tasks Reliability Scalability Each will be evaluated according to some variables: Number of robots Number of tasks Map area size Algorithm’s parameters Mohamed Gomaa Ghanem German University in Cairo Multi-robot Task Allocation Using Meta-heuristic Optimization
Introduction Multi-Robot Systems Task Allocation in MRS Proposed Approach Result & Discussion Conclusion & Future Work Evaluation Metrics 20/31 Cost of traveled distance Cost = numberOfRobots i=0 [dist(Ri , T0) + numOfTasks−1forRi j=0 dist(Tj, Tj+1)] Mission completion time = Algorithm Running time + Time taken by last robot to finish its tasks Reliability Scalability Each will be evaluated according to some variables: Number of robots Number of tasks Map area size Algorithm’s parameters Mohamed Gomaa Ghanem German University in Cairo Multi-robot Task Allocation Using Meta-heuristic Optimization
Introduction Multi-Robot Systems Task Allocation in MRS Proposed Approach Result & Discussion Conclusion & Future Work Evaluation Metrics 20/31 Cost of traveled distance Cost = numberOfRobots i=0 [dist(Ri , T0) + numOfTasks−1forRi j=0 dist(Tj, Tj+1)] Mission completion time = Algorithm Running time + Time taken by last robot to finish its tasks Reliability Scalability Each will be evaluated according to some variables: Number of robots Number of tasks Map area size Algorithm’s parameters Mohamed Gomaa Ghanem German University in Cairo Multi-robot Task Allocation Using Meta-heuristic Optimization
Introduction Multi-Robot Systems Task Allocation in MRS Proposed Approach Result & Discussion Conclusion & Future Work Evaluation Metrics 20/31 Cost of traveled distance Cost = numberOfRobots i=0 [dist(Ri , T0) + numOfTasks−1forRi j=0 dist(Tj, Tj+1)] Mission completion time = Algorithm Running time + Time taken by last robot to finish its tasks Reliability Scalability Each will be evaluated according to some variables: Number of robots Number of tasks Map area size Algorithm’s parameters Mohamed Gomaa Ghanem German University in Cairo Multi-robot Task Allocation Using Meta-heuristic Optimization
Introduction Multi-Robot Systems Task Allocation in MRS Proposed Approach Result & Discussion Conclusion & Future Work Evaluation Metrics 20/31 Cost of traveled distance Cost = numberOfRobots i=0 [dist(Ri , T0) + numOfTasks−1forRi j=0 dist(Tj, Tj+1)] Mission completion time = Algorithm Running time + Time taken by last robot to finish its tasks Reliability Scalability Each will be evaluated according to some variables: Number of robots Number of tasks Map area size Algorithm’s parameters Mohamed Gomaa Ghanem German University in Cairo Multi-robot Task Allocation Using Meta-heuristic Optimization
Introduction Multi-Robot Systems Task Allocation in MRS Proposed Approach Result & Discussion Conclusion & Future Work Evaluation Metrics 20/31 Cost of traveled distance Cost = numberOfRobots i=0 [dist(Ri , T0) + numOfTasks−1forRi j=0 dist(Tj, Tj+1)] Mission completion time = Algorithm Running time + Time taken by last robot to finish its tasks Reliability Scalability Each will be evaluated according to some variables: Number of robots Number of tasks Map area size Algorithm’s parameters Mohamed Gomaa Ghanem German University in Cairo Multi-robot Task Allocation Using Meta-heuristic Optimization
Introduction Multi-Robot Systems Task Allocation in MRS Proposed Approach Result & Discussion Conclusion & Future Work Evaluation Metrics 20/31 Cost of traveled distance Cost = numberOfRobots i=0 [dist(Ri , T0) + numOfTasks−1forRi j=0 dist(Tj, Tj+1)] Mission completion time = Algorithm Running time + Time taken by last robot to finish its tasks Reliability Scalability Each will be evaluated according to some variables: Number of robots Number of tasks Map area size Algorithm’s parameters Mohamed Gomaa Ghanem German University in Cairo Multi-robot Task Allocation Using Meta-heuristic Optimization
Introduction Multi-Robot Systems Task Allocation in MRS Proposed Approach Result & Discussion Conclusion & Future Work Java4MRS Application Mechanism 21/31 Mohamed Gomaa Ghanem German University in Cairo Multi-robot Task Allocation Using Meta-heuristic Optimization
Introduction Multi-Robot Systems Task Allocation in MRS Proposed Approach Result & Discussion Conclusion & Future Work Results 22/31 Parameters Tunning Tabu Search Number of iterations Nt ∗ Nr Tabu length 2 √ Nr + Nt Candidates per Iteration max(Nt,Nr ) Genetic Algorithm Number of iterations more is better till (Nr + Nt)2 Population Size more is better till (Nr + Nt)2 Swap Mutation Probability according to the problem(0.9) Mohamed Gomaa Ghanem German University in Cairo Multi-robot Task Allocation Using Meta-heuristic Optimization
Introduction Multi-Robot Systems Task Allocation in MRS Proposed Approach Result & Discussion Conclusion & Future Work Results 22/31 Parameters Tunning Tabu Search Number of iterations Nt ∗ Nr Tabu length 2 √ Nr + Nt Candidates per Iteration max(Nt,Nr ) Genetic Algorithm Number of iterations more is better till (Nr + Nt)2 Population Size more is better till (Nr + Nt)2 Swap Mutation Probability according to the problem(0.9) Mohamed Gomaa Ghanem German University in Cairo Multi-robot Task Allocation Using Meta-heuristic Optimization
Introduction Multi-Robot Systems Task Allocation in MRS Proposed Approach Result & Discussion Conclusion & Future Work Results 22/31 Parameters Tunning Tabu Search Number of iterations Nt ∗ Nr Tabu length 2 √ Nr + Nt Candidates per Iteration max(Nt,Nr ) Genetic Algorithm Number of iterations more is better till (Nr + Nt)2 Population Size more is better till (Nr + Nt)2 Swap Mutation Probability according to the problem(0.9) Mohamed Gomaa Ghanem German University in Cairo Multi-robot Task Allocation Using Meta-heuristic Optimization
Introduction Multi-Robot Systems Task Allocation in MRS Proposed Approach Result & Discussion Conclusion & Future Work Results 22/31 Parameters Tunning Tabu Search Number of iterations Nt ∗ Nr Tabu length 2 √ Nr + Nt Candidates per Iteration max(Nt,Nr ) Genetic Algorithm Number of iterations more is better till (Nr + Nt)2 Population Size more is better till (Nr + Nt)2 Swap Mutation Probability according to the problem(0.9) Mohamed Gomaa Ghanem German University in Cairo Multi-robot Task Allocation Using Meta-heuristic Optimization
Introduction Multi-Robot Systems Task Allocation in MRS Proposed Approach Result & Discussion Conclusion & Future Work Results 22/31 Parameters Tunning Tabu Search Number of iterations Nt ∗ Nr Tabu length 2 √ Nr + Nt Candidates per Iteration max(Nt,Nr ) Genetic Algorithm Number of iterations more is better till (Nr + Nt)2 Population Size more is better till (Nr + Nt)2 Swap Mutation Probability according to the problem(0.9) Mohamed Gomaa Ghanem German University in Cairo Multi-robot Task Allocation Using Meta-heuristic Optimization
Introduction Multi-Robot Systems Task Allocation in MRS Proposed Approach Result & Discussion Conclusion & Future Work Results 22/31 Parameters Tunning Tabu Search Number of iterations Nt ∗ Nr Tabu length 2 √ Nr + Nt Candidates per Iteration max(Nt,Nr ) Genetic Algorithm Number of iterations more is better till (Nr + Nt)2 Population Size more is better till (Nr + Nt)2 Swap Mutation Probability according to the problem(0.9) Mohamed Gomaa Ghanem German University in Cairo Multi-robot Task Allocation Using Meta-heuristic Optimization
Introduction Multi-Robot Systems Task Allocation in MRS Proposed Approach Result & Discussion Conclusion & Future Work Results 22/31 Parameters Tunning Tabu Search Number of iterations Nt ∗ Nr Tabu length 2 √ Nr + Nt Candidates per Iteration max(Nt,Nr ) Genetic Algorithm Number of iterations more is better till (Nr + Nt)2 Population Size more is better till (Nr + Nt)2 Swap Mutation Probability according to the problem(0.9) Mohamed Gomaa Ghanem German University in Cairo Multi-robot Task Allocation Using Meta-heuristic Optimization
Introduction Multi-Robot Systems Task Allocation in MRS Proposed Approach Result & Discussion Conclusion & Future Work Results 22/31 Parameters Tunning Tabu Search Number of iterations Nt ∗ Nr Tabu length 2 √ Nr + Nt Candidates per Iteration max(Nt,Nr ) Genetic Algorithm Number of iterations more is better till (Nr + Nt)2 Population Size more is better till (Nr + Nt)2 Swap Mutation Probability according to the problem(0.9) Mohamed Gomaa Ghanem German University in Cairo Multi-robot Task Allocation Using Meta-heuristic Optimization
Introduction Multi-Robot Systems Task Allocation in MRS Proposed Approach Result & Discussion Conclusion & Future Work Results 22/31 Parameters Tunning Tabu Search Number of iterations Nt ∗ Nr Tabu length 2 √ Nr + Nt Candidates per Iteration max(Nt,Nr ) Genetic Algorithm Number of iterations more is better till (Nr + Nt)2 Population Size more is better till (Nr + Nt)2 Swap Mutation Probability according to the problem(0.9) Mohamed Gomaa Ghanem German University in Cairo Multi-robot Task Allocation Using Meta-heuristic Optimization
Introduction Multi-Robot Systems Task Allocation in MRS Proposed Approach Result & Discussion Conclusion & Future Work Results 22/31 Parameters Tunning Tabu Search Number of iterations Nt ∗ Nr Tabu length 2 √ Nr + Nt Candidates per Iteration max(Nt,Nr ) Genetic Algorithm Number of iterations more is better till (Nr + Nt)2 Population Size more is better till (Nr + Nt)2 Swap Mutation Probability according to the problem(0.9) Mohamed Gomaa Ghanem German University in Cairo Multi-robot Task Allocation Using Meta-heuristic Optimization
Introduction Multi-Robot Systems Task Allocation in MRS Proposed Approach Result & Discussion Conclusion & Future Work Results 22/31 Parameters Tunning Tabu Search Number of iterations Nt ∗ Nr Tabu length 2 √ Nr + Nt Candidates per Iteration max(Nt,Nr ) Genetic Algorithm Number of iterations more is better till (Nr + Nt)2 Population Size more is better till (Nr + Nt)2 Swap Mutation Probability according to the problem(0.9) Mohamed Gomaa Ghanem German University in Cairo Multi-robot Task Allocation Using Meta-heuristic Optimization
Introduction Multi-Robot Systems Task Allocation in MRS Proposed Approach Result & Discussion Conclusion & Future Work Results 22/31 Parameters Tunning Tabu Search Number of iterations Nt ∗ Nr Tabu length 2 √ Nr + Nt Candidates per Iteration max(Nt,Nr ) Genetic Algorithm Number of iterations more is better till (Nr + Nt)2 Population Size more is better till (Nr + Nt)2 Swap Mutation Probability according to the problem(0.9) Mohamed Gomaa Ghanem German University in Cairo Multi-robot Task Allocation Using Meta-heuristic Optimization
Introduction Multi-Robot Systems Task Allocation in MRS Proposed Approach Result & Discussion Conclusion & Future Work Results 22/31 Parameters Tunning Tabu Search Number of iterations Nt ∗ Nr Tabu length 2 √ Nr + Nt Candidates per Iteration max(Nt,Nr ) Genetic Algorithm Number of iterations more is better till (Nr + Nt)2 Population Size more is better till (Nr + Nt)2 Swap Mutation Probability according to the problem(0.9) Mohamed Gomaa Ghanem German University in Cairo Multi-robot Task Allocation Using Meta-heuristic Optimization
Introduction Multi-Robot Systems Task Allocation in MRS Proposed Approach Result & Discussion Conclusion & Future Work Results 22/31 Parameters Tunning Tabu Search Number of iterations Nt ∗ Nr Tabu length 2 √ Nr + Nt Candidates per Iteration max(Nt,Nr ) Genetic Algorithm Number of iterations more is better till (Nr + Nt)2 Population Size more is better till (Nr + Nt)2 Swap Mutation Probability according to the problem(0.9) Mohamed Gomaa Ghanem German University in Cairo Multi-robot Task Allocation Using Meta-heuristic Optimization
Introduction Multi-Robot Systems Task Allocation in MRS Proposed Approach Result & Discussion Conclusion & Future Work Results 23/31 Number Of Tasks Mohamed Gomaa Ghanem German University in Cairo Multi-robot Task Allocation Using Meta-heuristic Optimization
Introduction Multi-Robot Systems Task Allocation in MRS Proposed Approach Result & Discussion Conclusion & Future Work Results 24/31 Number Of Tasks Mohamed Gomaa Ghanem German University in Cairo Multi-robot Task Allocation Using Meta-heuristic Optimization
Introduction Multi-Robot Systems Task Allocation in MRS Proposed Approach Result & Discussion Conclusion & Future Work Results 25/31 Number Of Tasks Mohamed Gomaa Ghanem German University in Cairo Multi-robot Task Allocation Using Meta-heuristic Optimization
Introduction Multi-Robot Systems Task Allocation in MRS Proposed Approach Result & Discussion Conclusion & Future Work Results 26/31 Number Of Robots Mohamed Gomaa Ghanem German University in Cairo Multi-robot Task Allocation Using Meta-heuristic Optimization
Introduction Multi-Robot Systems Task Allocation in MRS Proposed Approach Result & Discussion Conclusion & Future Work Results 27/31 Number Of Robots Mohamed Gomaa Ghanem German University in Cairo Multi-robot Task Allocation Using Meta-heuristic Optimization
Introduction Multi-Robot Systems Task Allocation in MRS Proposed Approach Result & Discussion Conclusion & Future Work Results 28/31 Number Of Robots Mohamed Gomaa Ghanem German University in Cairo Multi-robot Task Allocation Using Meta-heuristic Optimization
Introduction Multi-Robot Systems Task Allocation in MRS Proposed Approach Result & Discussion Conclusion & Future Work Conclusion 29/31 Experimental results showed that it’s better to use the Genetic Algorithm if an initial solution is not known, but if there exists an initial solution and trying to reach a better one therefor the tabu search outperforms genetic algorithm as it tries to improve the allocation in each iteration and to escape from a local minimum to a global minimum Tabu Search is also better in solving for real time as it provides faster response A hybrid approachs are examined and found that it’s better to get the initial solution from Genetic Algorithm and use it to start Tabu search in order to find out if there exist a better near solutions Mohamed Gomaa Ghanem German University in Cairo Multi-robot Task Allocation Using Meta-heuristic Optimization
Introduction Multi-Robot Systems Task Allocation in MRS Proposed Approach Result & Discussion Conclusion & Future Work Conclusion 29/31 Experimental results showed that it’s better to use the Genetic Algorithm if an initial solution is not known, but if there exists an initial solution and trying to reach a better one therefor the tabu search outperforms genetic algorithm as it tries to improve the allocation in each iteration and to escape from a local minimum to a global minimum Tabu Search is also better in solving for real time as it provides faster response A hybrid approachs are examined and found that it’s better to get the initial solution from Genetic Algorithm and use it to start Tabu search in order to find out if there exist a better near solutions Mohamed Gomaa Ghanem German University in Cairo Multi-robot Task Allocation Using Meta-heuristic Optimization
Introduction Multi-Robot Systems Task Allocation in MRS Proposed Approach Result & Discussion Conclusion & Future Work Conclusion 29/31 Experimental results showed that it’s better to use the Genetic Algorithm if an initial solution is not known, but if there exists an initial solution and trying to reach a better one therefor the tabu search outperforms genetic algorithm as it tries to improve the allocation in each iteration and to escape from a local minimum to a global minimum Tabu Search is also better in solving for real time as it provides faster response A hybrid approachs are examined and found that it’s better to get the initial solution from Genetic Algorithm and use it to start Tabu search in order to find out if there exist a better near solutions Mohamed Gomaa Ghanem German University in Cairo Multi-robot Task Allocation Using Meta-heuristic Optimization
Introduction Multi-Robot Systems Task Allocation in MRS Proposed Approach Result & Discussion Conclusion & Future Work Conclusion 29/31 Experimental results showed that it’s better to use the Genetic Algorithm if an initial solution is not known, but if there exists an initial solution and trying to reach a better one therefor the tabu search outperforms genetic algorithm as it tries to improve the allocation in each iteration and to escape from a local minimum to a global minimum Tabu Search is also better in solving for real time as it provides faster response A hybrid approachs are examined and found that it’s better to get the initial solution from Genetic Algorithm and use it to start Tabu search in order to find out if there exist a better near solutions Mohamed Gomaa Ghanem German University in Cairo Multi-robot Task Allocation Using Meta-heuristic Optimization
Introduction Multi-Robot Systems Task Allocation in MRS Proposed Approach Result & Discussion Conclusion & Future Work Future Work 30/31 Robots and Tasks heterogeneity More Robot Planning (obstical avoidance) Integrate Google Map Input ROS(Robot Operating System) insted of Simbad-3D Mohamed Gomaa Ghanem German University in Cairo Multi-robot Task Allocation Using Meta-heuristic Optimization
Introduction Multi-Robot Systems Task Allocation in MRS Proposed Approach Result & Discussion Conclusion & Future Work Future Work 30/31 Robots and Tasks heterogeneity More Robot Planning (obstical avoidance) Integrate Google Map Input ROS(Robot Operating System) insted of Simbad-3D Mohamed Gomaa Ghanem German University in Cairo Multi-robot Task Allocation Using Meta-heuristic Optimization
Introduction Multi-Robot Systems Task Allocation in MRS Proposed Approach Result & Discussion Conclusion & Future Work Future Work 30/31 Robots and Tasks heterogeneity More Robot Planning (obstical avoidance) Integrate Google Map Input ROS(Robot Operating System) insted of Simbad-3D Mohamed Gomaa Ghanem German University in Cairo Multi-robot Task Allocation Using Meta-heuristic Optimization
Introduction Multi-Robot Systems Task Allocation in MRS Proposed Approach Result & Discussion Conclusion & Future Work Future Work 30/31 Robots and Tasks heterogeneity More Robot Planning (obstical avoidance) Integrate Google Map Input ROS(Robot Operating System) insted of Simbad-3D Mohamed Gomaa Ghanem German University in Cairo Multi-robot Task Allocation Using Meta-heuristic Optimization
Introduction Multi-Robot Systems Task Allocation in MRS Proposed Approach Result & Discussion Conclusion & Future Work Future Work 30/31 Robots and Tasks heterogeneity More Robot Planning (obstical avoidance) Integrate Google Map Input ROS(Robot Operating System) insted of Simbad-3D Mohamed Gomaa Ghanem German University in Cairo Multi-robot Task Allocation Using Meta-heuristic Optimization
Introduction Multi-Robot Systems Task Allocation in MRS Proposed Approach Result & Discussion Conclusion & Future Work 1. download from : https://javar.googlecode.com/files/Java4Robot.jar 2. run & enjoy ... Demo Mohamed Gomaa Ghanem German University in Cairo Multi-robot Task Allocation Using Meta-heuristic Optimization

More Related Content

What's hot (11)

PPT
LFSR
CVR college of Engineering
 
PPTX
I2C Protocol
Sudhanshu Janwadkar
 
PPTX
Fake News Detection Using Machine learning algorithm
MudasirBashir23
 
PPT
microprocessor
ATTO RATHORE
 
PDF
Fraud detection with Machine Learning
Scaleway
 
PPT
Embedded system design process_models
Ravi Selvaraj
 
PPTX
Lyme disease
Heba Al-Thuwaini
 
PPTX
Mycobacterium tuberculosis
Awaaz Batazoo
 
PPTX
Day 1 (Lecture 3): Predictive Analytics in Healthcare
Aseda Owusua Addai-Deseh
 
PPT
Biometric's final ppt
Ankita Vanage
 
PPT
06. thumb instructions
balaji raja rajan Venkatachalam
 
LFSR
CVR college of Engineering
 
I2C Protocol
Sudhanshu Janwadkar
 
Fake News Detection Using Machine learning algorithm
MudasirBashir23
 
microprocessor
ATTO RATHORE
 
Fraud detection with Machine Learning
Scaleway
 
Embedded system design process_models
Ravi Selvaraj
 
Lyme disease
Heba Al-Thuwaini
 
Mycobacterium tuberculosis
Awaaz Batazoo
 
Day 1 (Lecture 3): Predictive Analytics in Healthcare
Aseda Owusua Addai-Deseh
 
Biometric's final ppt
Ankita Vanage
 
06. thumb instructions
balaji raja rajan Venkatachalam
 

Viewers also liked (20)

PDF
A Distributed Architecture for Heterogeneous Multi-Sensor Task Allocation
Diego Pizzocaro
 
PDF
On Multi-Sensor Task Allocation
Diego Pizzocaro
 
PDF
Improving Resource Utilization in Cloud using Application Placement Heuristics
AtakanAral
 
PPTX
Managing Innovation_Introduction to Innovation
OpenLearningLab
 
PPTX
Fan.xue.viva
DeWolf Xue
 
PPTX
Navid Conference - 1394 - Shiraz, Iran
Hamed Takmil
 
PDF
A genetic algorithm approach for multi objective optimization of supply chain...
Phuong Dx
 
PPTX
Top 50 java ee 7 best practices [con5669]
Ryan Cuprak
 
PDF
SDPM - Lecture 4 - Activity planning and resource allocation
OpenLearningLab
 
PPTX
50 EJB 3 Best Practices in 50 Minutes - JavaOne 2014
Ryan Cuprak
 
PPTX
Off page seo-سئو-کسب و کار الکترونیکی
Hanieh Ghofrani
 
PPTX
M phil thesis presentation
Dr Muhammad Mustansar
 
PDF
Multiobjective optimization and trade offs using pareto optimality
Amogh Mundhekar
 
PPTX
Pert,cpm, resource allocation and gert
Raj J Das
 
PPTX
Genetic Algorithm
SHIMI S L
 
PPT
Planning, scheduling and resource allocation
Jatin Mandhyan
 
PPT
My Mphil Power Point Presentation
Spices Board
 
PPTX
Genetic
saeedeh ezzati
 
PDF
My Thesis Defense Presentation
David Onoue
 
PDF
Making the most of your constrained resources optimizing resource allocation ...
Productivity Intelligence Institute
 
A Distributed Architecture for Heterogeneous Multi-Sensor Task Allocation
Diego Pizzocaro
 
On Multi-Sensor Task Allocation
Diego Pizzocaro
 
Improving Resource Utilization in Cloud using Application Placement Heuristics
AtakanAral
 
Managing Innovation_Introduction to Innovation
OpenLearningLab
 
Fan.xue.viva
DeWolf Xue
 
Navid Conference - 1394 - Shiraz, Iran
Hamed Takmil
 
A genetic algorithm approach for multi objective optimization of supply chain...
Phuong Dx
 
Top 50 java ee 7 best practices [con5669]
Ryan Cuprak
 
SDPM - Lecture 4 - Activity planning and resource allocation
OpenLearningLab
 
50 EJB 3 Best Practices in 50 Minutes - JavaOne 2014
Ryan Cuprak
 
Off page seo-سئو-کسب و کار الکترونیکی
Hanieh Ghofrani
 
M phil thesis presentation
Dr Muhammad Mustansar
 
Multiobjective optimization and trade offs using pareto optimality
Amogh Mundhekar
 
Pert,cpm, resource allocation and gert
Raj J Das
 
Genetic Algorithm
SHIMI S L
 
Planning, scheduling and resource allocation
Jatin Mandhyan
 
My Mphil Power Point Presentation
Spices Board
 
Genetic
saeedeh ezzati
 
My Thesis Defense Presentation
David Onoue
 
Making the most of your constrained resources optimizing resource allocation ...
Productivity Intelligence Institute
 
Ad

Similar to Multi-Robot Task Allocation using Meta-heuristic Optimization (20)

PPTX
Pareto-Optimal Search-Based Software Engineering (POSBSE): A Literature Survey
Abdel Salam Sayyad
 
DOCX
Mc0084 software project management & quality
smumbahelp
 
PDF
Operations Research Digital Material.pdf
TANVEERSINGHSOLANKI
 
PDF
Executable Metamodeling for Model V&V (May 25th, 2010)
Benoit Combemale
 
PDF
Introduction to Model-Based Machine Learning
Daniel Emaasit
 
PPTX
the application of machine lerning algorithm for SEE
KiranKumar671235
 
PDF
C013141723
IOSR Journals
 
PDF
Physical and Physic-Chemical Based Optimization Methods: A Review
Journal of Soft Computing in Civil Engineering
 
PDF
A MODEL TO COMPARE THE DEGREE OF REFACTORING OPPORTUNITIES OF THREE PROJECTS ...
acijjournal
 
PDF
A Model To Compare The Degree Of Refactoring Opportunities Of Three Projects ...
acijjournal
 
PDF
MultiObjective(11) - Copy
AMIT KUMAR
 
PDF
Dj4201737746
IJERA Editor
 
PPTX
Kernel based swarm optimization for renewable energy application
Aboul Ella Hassanien
 
PDF
CV2015
Maxim Sviridenko
 
PDF
An Integrated Solver For Optimization Problems
Monica Waters
 
PDF
Transcription Factor DNA Binding Prediction
UT, San Antonio
 
PDF
Mech vii-operation research [06 me74]-notes
Mallikarjunaswamy Swamy
 
PDF
An Adjacent Analysis of the Parallel Programming Model Perspective: A Survey
IRJET Journal
 
PDF
Multiobjective Firefly Algorithm for Continuous Optimization
Xin-She Yang
 
PDF
Application of Genetic Algorithm in Software Engineering: A Review
IRJESJOURNAL
 
Pareto-Optimal Search-Based Software Engineering (POSBSE): A Literature Survey
Abdel Salam Sayyad
 
Mc0084 software project management & quality
smumbahelp
 
Operations Research Digital Material.pdf
TANVEERSINGHSOLANKI
 
Executable Metamodeling for Model V&V (May 25th, 2010)
Benoit Combemale
 
Introduction to Model-Based Machine Learning
Daniel Emaasit
 
the application of machine lerning algorithm for SEE
KiranKumar671235
 
C013141723
IOSR Journals
 
Physical and Physic-Chemical Based Optimization Methods: A Review
Journal of Soft Computing in Civil Engineering
 
A MODEL TO COMPARE THE DEGREE OF REFACTORING OPPORTUNITIES OF THREE PROJECTS ...
acijjournal
 
A Model To Compare The Degree Of Refactoring Opportunities Of Three Projects ...
acijjournal
 
MultiObjective(11) - Copy
AMIT KUMAR
 
Dj4201737746
IJERA Editor
 
Kernel based swarm optimization for renewable energy application
Aboul Ella Hassanien
 
CV2015
Maxim Sviridenko
 
An Integrated Solver For Optimization Problems
Monica Waters
 
Transcription Factor DNA Binding Prediction
UT, San Antonio
 
Mech vii-operation research [06 me74]-notes
Mallikarjunaswamy Swamy
 
An Adjacent Analysis of the Parallel Programming Model Perspective: A Survey
IRJET Journal
 
Multiobjective Firefly Algorithm for Continuous Optimization
Xin-She Yang
 
Application of Genetic Algorithm in Software Engineering: A Review
IRJESJOURNAL
 
Ad

Recently uploaded (20)

PPTX
Electron Beam Machining for Production Process
Rajshahi University of Engineering & Technology(RUET), Bangladesh
 
PPTX
Introduction to Neural Networks and Perceptron Learning Algorithm.pptx
Kayalvizhi A
 
PDF
Additional Information in midterm CPE024 (1).pdf
abolisojoy
 
PDF
MOBILE AND WEB BASED REMOTE BUSINESS MONITORING SYSTEM
ijait
 
PDF
International Journal of Information Technology Convergence and services (IJI...
ijitcsjournal4
 
PDF
ARC--BUILDING-UTILITIES-2-PART-2 (1).pdf
IzzyBaniquedBusto
 
PDF
PORTFOLIO Golam Kibria Khan — architect with a passion for thoughtful design...
MasumKhan59
 
PDF
Statistical Data Analysis Using SPSS Software
shrikrishna kesharwani
 
PPTX
Structural Functiona theory this important for the theorist
cagumaydanny26
 
PPTX
Thermal runway and thermal stability.pptx
godow93766
 
PPTX
REINFORCEMENT AS CONSTRUCTION MATERIALS.pptx
mohaiminulhaquesami
 
PPTX
MobileComputingMANET2023 MobileComputingMANET2023.pptx
masterfake98765
 
PDF
Set Relation Function Practice session 24.05.2025.pdf
DrStephenStrange4
 
PPTX
ISO/IEC JTC 1/WG 9 (MAR) Convenor Report
Kurata Takeshi
 
PPTX
drones for disaster prevention response.pptx
NawrasShatnawi1
 
PDF
Water Design_Manual_2005. KENYA FOR WASTER SUPPLY AND SEWERAGE
DancanNgutuku
 
PDF
monopile foundation seminar topic for civil engineering students
Ahina5
 
PDF
Ethics and Trustworthy AI in Healthcare – Governing Sensitive Data, Profiling...
AlqualsaDIResearchGr
 
PPTX
Heart Bleed Bug - A case study (Course: Cryptography and Network Security)
Adri Jovin
 
PPTX
265587293-NFPA 101 Life safety code-PPT-1.pptx
chandermwason
 
Electron Beam Machining for Production Process
Rajshahi University of Engineering & Technology(RUET), Bangladesh
 
Introduction to Neural Networks and Perceptron Learning Algorithm.pptx
Kayalvizhi A
 
Additional Information in midterm CPE024 (1).pdf
abolisojoy
 
MOBILE AND WEB BASED REMOTE BUSINESS MONITORING SYSTEM
ijait
 
International Journal of Information Technology Convergence and services (IJI...
ijitcsjournal4
 
ARC--BUILDING-UTILITIES-2-PART-2 (1).pdf
IzzyBaniquedBusto
 
PORTFOLIO Golam Kibria Khan — architect with a passion for thoughtful design...
MasumKhan59
 
Statistical Data Analysis Using SPSS Software
shrikrishna kesharwani
 
Structural Functiona theory this important for the theorist
cagumaydanny26
 
Thermal runway and thermal stability.pptx
godow93766
 
REINFORCEMENT AS CONSTRUCTION MATERIALS.pptx
mohaiminulhaquesami
 
MobileComputingMANET2023 MobileComputingMANET2023.pptx
masterfake98765
 
Set Relation Function Practice session 24.05.2025.pdf
DrStephenStrange4
 
ISO/IEC JTC 1/WG 9 (MAR) Convenor Report
Kurata Takeshi
 
drones for disaster prevention response.pptx
NawrasShatnawi1
 
Water Design_Manual_2005. KENYA FOR WASTER SUPPLY AND SEWERAGE
DancanNgutuku
 
monopile foundation seminar topic for civil engineering students
Ahina5
 
Ethics and Trustworthy AI in Healthcare – Governing Sensitive Data, Profiling...
AlqualsaDIResearchGr
 
Heart Bleed Bug - A case study (Course: Cryptography and Network Security)
Adri Jovin
 
265587293-NFPA 101 Life safety code-PPT-1.pptx
chandermwason
 

Multi-Robot Task Allocation using Meta-heuristic Optimization

  • 1. Introduction Multi-Robot Systems Task Allocation in MRS Proposed Approach Result & Discussion Conclusion & Future Work Multi-robot Task Allocation Using Meta-heuristic Optimization Mohamed Gomaa Ghanem Supervised by: Dr. Eng. Alaa Khamis German University in Cairo June 25, 2012 Mohamed Gomaa Ghanem German University in Cairo Multi-robot Task Allocation Using Meta-heuristic Optimization
  • 2. Introduction Multi-Robot Systems Task Allocation in MRS Proposed Approach Result & Discussion Conclusion & Future Work Outline 1. Introduction Motivation Objective 2. Multi-robot System (MRS) Definition Application Challenging Aspects 3. Task Allocation in MRS Problem Formulation Architecture Approaches 4. Proposed Approach Meta-heuristic optimization techniques Tabu Search based Task Allocation Genetic Algorithm based Task Allocation Hybrid Approaches 5. Result & Discussion Experemental Setup Evaluation Metrics Results 6. Conclusion & Future Work Conclusion Future work Mohamed Gomaa Ghanem German University in Cairo Multi-robot Task Allocation Using Meta-heuristic Optimization
  • 3. Introduction Multi-Robot Systems Task Allocation in MRS Proposed Approach Result & Discussion Conclusion & Future Work Outline 1. Introduction Motivation Objective 2. Multi-robot System (MRS) Definition Application Challenging Aspects 3. Task Allocation in MRS Problem Formulation Architecture Approaches 4. Proposed Approach Meta-heuristic optimization techniques Tabu Search based Task Allocation Genetic Algorithm based Task Allocation Hybrid Approaches 5. Result & Discussion Experemental Setup Evaluation Metrics Results 6. Conclusion & Future Work Conclusion Future work Mohamed Gomaa Ghanem German University in Cairo Multi-robot Task Allocation Using Meta-heuristic Optimization
  • 4. Introduction Multi-Robot Systems Task Allocation in MRS Proposed Approach Result & Discussion Conclusion & Future Work Outline 1. Introduction Motivation Objective 2. Multi-robot System (MRS) Definition Application Challenging Aspects 3. Task Allocation in MRS Problem Formulation Architecture Approaches 4. Proposed Approach Meta-heuristic optimization techniques Tabu Search based Task Allocation Genetic Algorithm based Task Allocation Hybrid Approaches 5. Result & Discussion Experemental Setup Evaluation Metrics Results 6. Conclusion & Future Work Conclusion Future work Mohamed Gomaa Ghanem German University in Cairo Multi-robot Task Allocation Using Meta-heuristic Optimization
  • 5. Introduction Multi-Robot Systems Task Allocation in MRS Proposed Approach Result & Discussion Conclusion & Future Work Outline 1. Introduction Motivation Objective 2. Multi-robot System (MRS) Definition Application Challenging Aspects 3. Task Allocation in MRS Problem Formulation Architecture Approaches 4. Proposed Approach Meta-heuristic optimization techniques Tabu Search based Task Allocation Genetic Algorithm based Task Allocation Hybrid Approaches 5. Result & Discussion Experemental Setup Evaluation Metrics Results 6. Conclusion & Future Work Conclusion Future work Mohamed Gomaa Ghanem German University in Cairo Multi-robot Task Allocation Using Meta-heuristic Optimization
  • 6. Introduction Multi-Robot Systems Task Allocation in MRS Proposed Approach Result & Discussion Conclusion & Future Work Outline 1. Introduction Motivation Objective 2. Multi-robot System (MRS) Definition Application Challenging Aspects 3. Task Allocation in MRS Problem Formulation Architecture Approaches 4. Proposed Approach Meta-heuristic optimization techniques Tabu Search based Task Allocation Genetic Algorithm based Task Allocation Hybrid Approaches 5. Result & Discussion Experemental Setup Evaluation Metrics Results 6. Conclusion & Future Work Conclusion Future work Mohamed Gomaa Ghanem German University in Cairo Multi-robot Task Allocation Using Meta-heuristic Optimization
  • 7. Introduction Multi-Robot Systems Task Allocation in MRS Proposed Approach Result & Discussion Conclusion & Future Work Outline 1. Introduction Motivation Objective 2. Multi-robot System (MRS) Definition Application Challenging Aspects 3. Task Allocation in MRS Problem Formulation Architecture Approaches 4. Proposed Approach Meta-heuristic optimization techniques Tabu Search based Task Allocation Genetic Algorithm based Task Allocation Hybrid Approaches 5. Result & Discussion Experemental Setup Evaluation Metrics Results 6. Conclusion & Future Work Conclusion Future work Mohamed Gomaa Ghanem German University in Cairo Multi-robot Task Allocation Using Meta-heuristic Optimization
  • 8. Introduction Multi-Robot Systems Task Allocation in MRS Proposed Approach Result & Discussion Conclusion & Future Work Motivation 3/31 Who will do What by When? Mohamed Gomaa Ghanem German University in Cairo Multi-robot Task Allocation Using Meta-heuristic Optimization
  • 9. Introduction Multi-Robot Systems Task Allocation in MRS Proposed Approach Result & Discussion Conclusion & Future Work Motivation 3/31 Who will do What by When? Mohamed Gomaa Ghanem German University in Cairo Multi-robot Task Allocation Using Meta-heuristic Optimization
  • 10. Introduction Multi-Robot Systems Task Allocation in MRS Proposed Approach Result & Discussion Conclusion & Future Work Objective 4/31 Project Objectives Finding out which meta-heuristic is better in solving task allocation problem Implementing a framework that allows comparing and simulating a running algorithm(s) Mohamed Gomaa Ghanem German University in Cairo Multi-robot Task Allocation Using Meta-heuristic Optimization
  • 11. Introduction Multi-Robot Systems Task Allocation in MRS Proposed Approach Result & Discussion Conclusion & Future Work Objective 4/31 Project Objectives Finding out which meta-heuristic is better in solving task allocation problem Implementing a framework that allows comparing and simulating a running algorithm(s) Mohamed Gomaa Ghanem German University in Cairo Multi-robot Task Allocation Using Meta-heuristic Optimization
  • 12. Introduction Multi-Robot Systems Task Allocation in MRS Proposed Approach Result & Discussion Conclusion & Future Work Objective 4/31 Project Objectives Finding out which meta-heuristic is better in solving task allocation problem Implementing a framework that allows comparing and simulating a running algorithm(s) Mohamed Gomaa Ghanem German University in Cairo Multi-robot Task Allocation Using Meta-heuristic Optimization
  • 13. Introduction Multi-Robot Systems Task Allocation in MRS Proposed Approach Result & Discussion Conclusion & Future Work Diffenetion 5/31 Multi-robot Systems are a group of robots that are designed aiming to perform some collection behavior Multi-robot System do resolve task complexity, increase performance, have more reliability, simple in design, and easily to be managed through a host computer. Mohamed Gomaa Ghanem German University in Cairo Multi-robot Task Allocation Using Meta-heuristic Optimization
  • 14. Introduction Multi-Robot Systems Task Allocation in MRS Proposed Approach Result & Discussion Conclusion & Future Work Diffenetion 5/31 Multi-robot Systems are a group of robots that are designed aiming to perform some collection behavior Multi-robot System do resolve task complexity, increase performance, have more reliability, simple in design, and easily to be managed through a host computer. Mohamed Gomaa Ghanem German University in Cairo Multi-robot Task Allocation Using Meta-heuristic Optimization
  • 15. Introduction Multi-Robot Systems Task Allocation in MRS Proposed Approach Result & Discussion Conclusion & Future Work Applications 6/31 The application of MRS includes, but is not limited to : Autonomous inspection of complex engineered structures Distributed sensing tasks in micro machinery or the human body Killing Cancer Tumors in Human Body Mining Agricultural Foraging Cooperative Tracking Surveillance, Reconnaissance and Intelligence Mohamed Gomaa Ghanem German University in Cairo Multi-robot Task Allocation Using Meta-heuristic Optimization
  • 16. Introduction Multi-Robot Systems Task Allocation in MRS Proposed Approach Result & Discussion Conclusion & Future Work Challenging Aspects 7/31 MRS has many challenging aspects that represent an open research problems that could differ from one problem to another, Some of common challenging problem of MRS are: Analysis and Modeling the Problem Algorithm Design Implementation and Test Mohamed Gomaa Ghanem German University in Cairo Multi-robot Task Allocation Using Meta-heuristic Optimization
  • 17. Introduction Multi-Robot Systems Task Allocation in MRS Proposed Approach Result & Discussion Conclusion & Future Work Challenging Aspects 7/31 MRS has many challenging aspects that represent an open research problems that could differ from one problem to another, Some of common challenging problem of MRS are: Analysis and Modeling the Problem Algorithm Design Implementation and Test Mohamed Gomaa Ghanem German University in Cairo Multi-robot Task Allocation Using Meta-heuristic Optimization
  • 18. Introduction Multi-Robot Systems Task Allocation in MRS Proposed Approach Result & Discussion Conclusion & Future Work Challenging Aspects 7/31 MRS has many challenging aspects that represent an open research problems that could differ from one problem to another, Some of common challenging problem of MRS are: Analysis and Modeling the Problem Algorithm Design Implementation and Test Mohamed Gomaa Ghanem German University in Cairo Multi-robot Task Allocation Using Meta-heuristic Optimization
  • 19. Introduction Multi-Robot Systems Task Allocation in MRS Proposed Approach Result & Discussion Conclusion & Future Work Challenging Aspects 7/31 MRS has many challenging aspects that represent an open research problems that could differ from one problem to another, Some of common challenging problem of MRS are: Analysis and Modeling the Problem Algorithm Design Implementation and Test Mohamed Gomaa Ghanem German University in Cairo Multi-robot Task Allocation Using Meta-heuristic Optimization
  • 20. Introduction Multi-Robot Systems Task Allocation in MRS Proposed Approach Result & Discussion Conclusion & Future Work Challenging Aspects 7/31 MRS has many challenging aspects that represent an open research problems that could differ from one problem to another, Some of common challenging problem of MRS are: Analysis and Modeling the Problem Algorithm Design Implementation and Test Mohamed Gomaa Ghanem German University in Cairo Multi-robot Task Allocation Using Meta-heuristic Optimization
  • 21. Introduction Multi-Robot Systems Task Allocation in MRS Proposed Approach Result & Discussion Conclusion & Future Work Problem Formulation 8/31 Task allocation is a NP hard problem that addresses how to optimally assign a set of tasks to a set of robots to maximize overall expected performance, taking into account the priorities of the tasks and the skill ratings of the robots. The goal is to find out which algorithm is better in allocating N number of robots to M number of tasks laying on an area of A2 and reaching the minimum Travel Distance and/or Mission Completion Time for all robots to cover all required tasks. Mohamed Gomaa Ghanem German University in Cairo Multi-robot Task Allocation Using Meta-heuristic Optimization
  • 22. Introduction Multi-Robot Systems Task Allocation in MRS Proposed Approach Result & Discussion Conclusion & Future Work Problem Formulation 8/31 Task allocation is a NP hard problem that addresses how to optimally assign a set of tasks to a set of robots to maximize overall expected performance, taking into account the priorities of the tasks and the skill ratings of the robots. The goal is to find out which algorithm is better in allocating N number of robots to M number of tasks laying on an area of A2 and reaching the minimum Travel Distance and/or Mission Completion Time for all robots to cover all required tasks. Mohamed Gomaa Ghanem German University in Cairo Multi-robot Task Allocation Using Meta-heuristic Optimization
  • 23. Introduction Multi-Robot Systems Task Allocation in MRS Proposed Approach Result & Discussion Conclusion & Future Work Architecture Centralized Architecture Robot team treated as a single ”system” with many degrees of freedom. A single robot is the ”leader”, which plans optimal actions for group. Group members send information to leader and carry out actions. Pros Cons Leader can take all relevant information into ac- count. Computationally hard and sometimes intractable for more than a few robots. In theory, coordination can be perfect: Optimal plans possible. Makes unrealistic assumptions, where all relevant info can be transmitted to leader, and this info doesn’t change during plan construction. Result: response sluggish or inaccurate Vulnerable to malfunction of leader Heavy communication load Mohamed Gomaa Ghanem German University in Cairo Multi-robot Task Allocation Using Meta-heuristic Optimization
  • 24. Introduction Multi-Robot Systems Task Allocation in MRS Proposed Approach Result & Discussion Conclusion & Future Work Architecture Centralized Architecture Robot team treated as a single ”system” with many degrees of freedom. A single robot is the ”leader”, which plans optimal actions for group. Group members send information to leader and carry out actions. Pros Cons Leader can take all relevant information into ac- count. Computationally hard and sometimes intractable for more than a few robots. In theory, coordination can be perfect: Optimal plans possible. Makes unrealistic assumptions, where all relevant info can be transmitted to leader, and this info doesn’t change during plan construction. Result: response sluggish or inaccurate Vulnerable to malfunction of leader Heavy communication load Mohamed Gomaa Ghanem German University in Cairo Multi-robot Task Allocation Using Meta-heuristic Optimization
  • 25. Introduction Multi-Robot Systems Task Allocation in MRS Proposed Approach Result & Discussion Conclusion & Future Work Architecture Decentralized Architecture Distributed architecture is concentrated in planning responsibility spread over team, where each robot basically independent from the others and robots use locally observable information to make their plans. Pros Cons Fast response to dynamic conditions. Not all problems can be decomposed well. Little or no communication required. Plans are based only on local information. Little computation required. Result: solutions are often highly suboptimal. Smooth response to environmental changes. Mohamed Gomaa Ghanem German University in Cairo Multi-robot Task Allocation Using Meta-heuristic Optimization
  • 26. Introduction Multi-Robot Systems Task Allocation in MRS Proposed Approach Result & Discussion Conclusion & Future Work Architecture Decentralized Architecture Distributed architecture is concentrated in planning responsibility spread over team, where each robot basically independent from the others and robots use locally observable information to make their plans. Pros Cons Fast response to dynamic conditions. Not all problems can be decomposed well. Little or no communication required. Plans are based only on local information. Little computation required. Result: solutions are often highly suboptimal. Smooth response to environmental changes. Mohamed Gomaa Ghanem German University in Cairo Multi-robot Task Allocation Using Meta-heuristic Optimization
  • 27. Introduction Multi-Robot Systems Task Allocation in MRS Proposed Approach Result & Discussion Conclusion & Future Work Approaches 11/31 Market Based Approaches Market-based approach is based on the economic model of a free market, each robot seeks to maximize individual ”profit”, Robots can negotiate and bid for tasks individual profit helps the common good, and decisions are made locally but effects approach optimality. Pros Cons Robustness and quickness of distributed sys- tem. Cost heuristics can be inaccurate. Approaches optimality of centralized sys- tem. Much of this approach is still under development. Low communication requirements. Mohamed Gomaa Ghanem German University in Cairo Multi-robot Task Allocation Using Meta-heuristic Optimization
  • 28. Introduction Multi-Robot Systems Task Allocation in MRS Proposed Approach Result & Discussion Conclusion & Future Work Approaches 11/31 Market Based Approaches Market-based approach is based on the economic model of a free market, each robot seeks to maximize individual ”profit”, Robots can negotiate and bid for tasks individual profit helps the common good, and decisions are made locally but effects approach optimality. Pros Cons Robustness and quickness of distributed sys- tem. Cost heuristics can be inaccurate. Approaches optimality of centralized sys- tem. Much of this approach is still under development. Low communication requirements. Mohamed Gomaa Ghanem German University in Cairo Multi-robot Task Allocation Using Meta-heuristic Optimization
  • 29. Introduction Multi-Robot Systems Task Allocation in MRS Proposed Approach Result & Discussion Conclusion & Future Work Approaches 12/31 Optimization Based Approaches Response Surface Methodology. Gradient-Based Search. Heuristic Searches. Mohamed Gomaa Ghanem German University in Cairo Multi-robot Task Allocation Using Meta-heuristic Optimization
  • 30. Introduction Multi-Robot Systems Task Allocation in MRS Proposed Approach Result & Discussion Conclusion & Future Work Approaches 12/31 Optimization Based Approaches Response Surface Methodology. Gradient-Based Search. Heuristic Searches. Mohamed Gomaa Ghanem German University in Cairo Multi-robot Task Allocation Using Meta-heuristic Optimization
  • 31. Introduction Multi-Robot Systems Task Allocation in MRS Proposed Approach Result & Discussion Conclusion & Future Work Approaches 12/31 Optimization Based Approaches Response Surface Methodology. Gradient-Based Search. Heuristic Searches. Mohamed Gomaa Ghanem German University in Cairo Multi-robot Task Allocation Using Meta-heuristic Optimization
  • 32. Introduction Multi-Robot Systems Task Allocation in MRS Proposed Approach Result & Discussion Conclusion & Future Work Approaches 12/31 Optimization Based Approaches Response Surface Methodology. Gradient-Based Search. Heuristic Searches. Mohamed Gomaa Ghanem German University in Cairo Multi-robot Task Allocation Using Meta-heuristic Optimization
  • 33. Introduction Multi-Robot Systems Task Allocation in MRS Proposed Approach Result & Discussion Conclusion & Future Work Meta-heuristic optimization techniques 13/31 Meta-heuristic classifications Main classification Used Algorithms Mohamed Gomaa Ghanem German University in Cairo Multi-robot Task Allocation Using Meta-heuristic Optimization
  • 34. Introduction Multi-Robot Systems Task Allocation in MRS Proposed Approach Result & Discussion Conclusion & Future Work Meta-heuristic optimization techniques 13/31 Meta-heuristic classifications Main classification Used Algorithms Mohamed Gomaa Ghanem German University in Cairo Multi-robot Task Allocation Using Meta-heuristic Optimization
  • 35. Introduction Multi-Robot Systems Task Allocation in MRS Proposed Approach Result & Discussion Conclusion & Future Work Meta-heuristic optimization techniques 13/31 Meta-heuristic classifications Main classification Used Algorithms Mohamed Gomaa Ghanem German University in Cairo Multi-robot Task Allocation Using Meta-heuristic Optimization
  • 36. Introduction Multi-Robot Systems Task Allocation in MRS Proposed Approach Result & Discussion Conclusion & Future Work Tabu Search based Task Allocation 14/31 TS Properties Trajectory based Local search With memory Naturally inspired Pros Cons Allow non-improving solution to accept in order to escape from local optimum. Global optimum may not be found, depends on parameter settings. Can be applied to both discrete and contin- uous solution spaces. Can obtain solutions that often surpass a best solution previously found by other ap- proaches. Mohamed Gomaa Ghanem German University in Cairo Multi-robot Task Allocation Using Meta-heuristic Optimization
  • 37. Introduction Multi-Robot Systems Task Allocation in MRS Proposed Approach Result & Discussion Conclusion & Future Work Tabu Search based Task Allocation 14/31 TS Properties Trajectory based Local search With memory Naturally inspired Pros Cons Allow non-improving solution to accept in order to escape from local optimum. Global optimum may not be found, depends on parameter settings. Can be applied to both discrete and contin- uous solution spaces. Can obtain solutions that often surpass a best solution previously found by other ap- proaches. Mohamed Gomaa Ghanem German University in Cairo Multi-robot Task Allocation Using Meta-heuristic Optimization
  • 38. Introduction Multi-Robot Systems Task Allocation in MRS Proposed Approach Result & Discussion Conclusion & Future Work Tabu Search based Task Allocation 14/31 TS Properties Trajectory based Local search With memory Naturally inspired Pros Cons Allow non-improving solution to accept in order to escape from local optimum. Global optimum may not be found, depends on parameter settings. Can be applied to both discrete and contin- uous solution spaces. Can obtain solutions that often surpass a best solution previously found by other ap- proaches. Mohamed Gomaa Ghanem German University in Cairo Multi-robot Task Allocation Using Meta-heuristic Optimization
  • 39. Introduction Multi-Robot Systems Task Allocation in MRS Proposed Approach Result & Discussion Conclusion & Future Work Tabu Search based Task Allocation 14/31 TS Properties Trajectory based Local search With memory Naturally inspired Pros Cons Allow non-improving solution to accept in order to escape from local optimum. Global optimum may not be found, depends on parameter settings. Can be applied to both discrete and contin- uous solution spaces. Can obtain solutions that often surpass a best solution previously found by other ap- proaches. Mohamed Gomaa Ghanem German University in Cairo Multi-robot Task Allocation Using Meta-heuristic Optimization
  • 40. Introduction Multi-Robot Systems Task Allocation in MRS Proposed Approach Result & Discussion Conclusion & Future Work Tabu Search based Task Allocation 14/31 TS Properties Trajectory based Local search With memory Naturally inspired Pros Cons Allow non-improving solution to accept in order to escape from local optimum. Global optimum may not be found, depends on parameter settings. Can be applied to both discrete and contin- uous solution spaces. Can obtain solutions that often surpass a best solution previously found by other ap- proaches. Mohamed Gomaa Ghanem German University in Cairo Multi-robot Task Allocation Using Meta-heuristic Optimization
  • 41. Introduction Multi-Robot Systems Task Allocation in MRS Proposed Approach Result & Discussion Conclusion & Future Work Tabu Search based Task Allocation 14/31 TS Properties Trajectory based Local search With memory Naturally inspired Pros Cons Allow non-improving solution to accept in order to escape from local optimum. Global optimum may not be found, depends on parameter settings. Can be applied to both discrete and contin- uous solution spaces. Can obtain solutions that often surpass a best solution previously found by other ap- proaches. Mohamed Gomaa Ghanem German University in Cairo Multi-robot Task Allocation Using Meta-heuristic Optimization
  • 42. Introduction Multi-Robot Systems Task Allocation in MRS Proposed Approach Result & Discussion Conclusion & Future Work Tabu Search based Task Allocation 15/31 TS Mechanism Mohamed Gomaa Ghanem German University in Cairo Multi-robot Task Allocation Using Meta-heuristic Optimization
  • 43. Introduction Multi-Robot Systems Task Allocation in MRS Proposed Approach Result & Discussion Conclusion & Future Work Genetic Algorithm based Task Allocation 16/31 GA Properties Population based With memory Naturally inspired Pros Cons Often locate good solutions. Time Delay. This is an effective heuristic when dealing with a very large solution space. Tend to converge towards local points, rather than global points Mutation introduces new information gene pool, that protects against converging too quickly to local optimum. Mohamed Gomaa Ghanem German University in Cairo Multi-robot Task Allocation Using Meta-heuristic Optimization
  • 44. Introduction Multi-Robot Systems Task Allocation in MRS Proposed Approach Result & Discussion Conclusion & Future Work Genetic Algorithm based Task Allocation 16/31 GA Properties Population based With memory Naturally inspired Pros Cons Often locate good solutions. Time Delay. This is an effective heuristic when dealing with a very large solution space. Tend to converge towards local points, rather than global points Mutation introduces new information gene pool, that protects against converging too quickly to local optimum. Mohamed Gomaa Ghanem German University in Cairo Multi-robot Task Allocation Using Meta-heuristic Optimization
  • 45. Introduction Multi-Robot Systems Task Allocation in MRS Proposed Approach Result & Discussion Conclusion & Future Work Genetic Algorithm based Task Allocation 16/31 GA Properties Population based With memory Naturally inspired Pros Cons Often locate good solutions. Time Delay. This is an effective heuristic when dealing with a very large solution space. Tend to converge towards local points, rather than global points Mutation introduces new information gene pool, that protects against converging too quickly to local optimum. Mohamed Gomaa Ghanem German University in Cairo Multi-robot Task Allocation Using Meta-heuristic Optimization
  • 46. Introduction Multi-Robot Systems Task Allocation in MRS Proposed Approach Result & Discussion Conclusion & Future Work Genetic Algorithm based Task Allocation 16/31 GA Properties Population based With memory Naturally inspired Pros Cons Often locate good solutions. Time Delay. This is an effective heuristic when dealing with a very large solution space. Tend to converge towards local points, rather than global points Mutation introduces new information gene pool, that protects against converging too quickly to local optimum. Mohamed Gomaa Ghanem German University in Cairo Multi-robot Task Allocation Using Meta-heuristic Optimization
  • 47. Introduction Multi-Robot Systems Task Allocation in MRS Proposed Approach Result & Discussion Conclusion & Future Work Genetic Algorithm based Task Allocation 16/31 GA Properties Population based With memory Naturally inspired Pros Cons Often locate good solutions. Time Delay. This is an effective heuristic when dealing with a very large solution space. Tend to converge towards local points, rather than global points Mutation introduces new information gene pool, that protects against converging too quickly to local optimum. Mohamed Gomaa Ghanem German University in Cairo Multi-robot Task Allocation Using Meta-heuristic Optimization
  • 48. Introduction Multi-Robot Systems Task Allocation in MRS Proposed Approach Result & Discussion Conclusion & Future Work Genetic Algorithm based Task Allocation 17/31 GA Mechanism Mohamed Gomaa Ghanem German University in Cairo Multi-robot Task Allocation Using Meta-heuristic Optimization
  • 49. Introduction Multi-Robot Systems Task Allocation in MRS Proposed Approach Result & Discussion Conclusion & Future Work Hybrid Approaches 18/31 Two Hybrid Approaches TS-GA:: best solution from TS to be considered as an intial solution in GA Require: IntialSolution & TasksPosition & RobotsPosition Solutionbest ← Tabu-Search(IntialSolution,TasksPosition,RobotsPosition) RUN Genetic-Algorithm(Solutionbest ,TasksPosition,RobotsPosition) return Best achieved solution GA-TS:: best solution from GA to be considered as an intial solution in TS Require: IntialSolution & TasksPosition & RobotsPosition Solutionbest ← Genetic-Algorithm(IntialSolution,TasksPosition,RobotsPosition) RUN Tabu-Search(Solutionbest ,TasksPosition,RobotsPosition) return Best achieved solution Mohamed Gomaa Ghanem German University in Cairo Multi-robot Task Allocation Using Meta-heuristic Optimization
  • 50. Introduction Multi-Robot Systems Task Allocation in MRS Proposed Approach Result & Discussion Conclusion & Future Work Hybrid Approaches 18/31 Two Hybrid Approaches TS-GA:: best solution from TS to be considered as an intial solution in GA Require: IntialSolution & TasksPosition & RobotsPosition Solutionbest ← Tabu-Search(IntialSolution,TasksPosition,RobotsPosition) RUN Genetic-Algorithm(Solutionbest ,TasksPosition,RobotsPosition) return Best achieved solution GA-TS:: best solution from GA to be considered as an intial solution in TS Require: IntialSolution & TasksPosition & RobotsPosition Solutionbest ← Genetic-Algorithm(IntialSolution,TasksPosition,RobotsPosition) RUN Tabu-Search(Solutionbest ,TasksPosition,RobotsPosition) return Best achieved solution Mohamed Gomaa Ghanem German University in Cairo Multi-robot Task Allocation Using Meta-heuristic Optimization
  • 51. Introduction Multi-Robot Systems Task Allocation in MRS Proposed Approach Result & Discussion Conclusion & Future Work Hybrid Approaches 18/31 Two Hybrid Approaches TS-GA:: best solution from TS to be considered as an intial solution in GA Require: IntialSolution & TasksPosition & RobotsPosition Solutionbest ← Tabu-Search(IntialSolution,TasksPosition,RobotsPosition) RUN Genetic-Algorithm(Solutionbest ,TasksPosition,RobotsPosition) return Best achieved solution GA-TS:: best solution from GA to be considered as an intial solution in TS Require: IntialSolution & TasksPosition & RobotsPosition Solutionbest ← Genetic-Algorithm(IntialSolution,TasksPosition,RobotsPosition) RUN Tabu-Search(Solutionbest ,TasksPosition,RobotsPosition) return Best achieved solution Mohamed Gomaa Ghanem German University in Cairo Multi-robot Task Allocation Using Meta-heuristic Optimization
  • 52. Introduction Multi-Robot Systems Task Allocation in MRS Proposed Approach Result & Discussion Conclusion & Future Work Hybrid Approaches 18/31 Two Hybrid Approaches TS-GA:: best solution from TS to be considered as an intial solution in GA Require: IntialSolution & TasksPosition & RobotsPosition Solutionbest ← Tabu-Search(IntialSolution,TasksPosition,RobotsPosition) RUN Genetic-Algorithm(Solutionbest ,TasksPosition,RobotsPosition) return Best achieved solution GA-TS:: best solution from GA to be considered as an intial solution in TS Require: IntialSolution & TasksPosition & RobotsPosition Solutionbest ← Genetic-Algorithm(IntialSolution,TasksPosition,RobotsPosition) RUN Tabu-Search(Solutionbest ,TasksPosition,RobotsPosition) return Best achieved solution Mohamed Gomaa Ghanem German University in Cairo Multi-robot Task Allocation Using Meta-heuristic Optimization
  • 53. Introduction Multi-Robot Systems Task Allocation in MRS Proposed Approach Result & Discussion Conclusion & Future Work Hybrid Approaches 18/31 Two Hybrid Approaches TS-GA:: best solution from TS to be considered as an intial solution in GA Require: IntialSolution & TasksPosition & RobotsPosition Solutionbest ← Tabu-Search(IntialSolution,TasksPosition,RobotsPosition) RUN Genetic-Algorithm(Solutionbest ,TasksPosition,RobotsPosition) return Best achieved solution GA-TS:: best solution from GA to be considered as an intial solution in TS Require: IntialSolution & TasksPosition & RobotsPosition Solutionbest ← Genetic-Algorithm(IntialSolution,TasksPosition,RobotsPosition) RUN Tabu-Search(Solutionbest ,TasksPosition,RobotsPosition) return Best achieved solution Mohamed Gomaa Ghanem German University in Cairo Multi-robot Task Allocation Using Meta-heuristic Optimization
  • 54. Introduction Multi-Robot Systems Task Allocation in MRS Proposed Approach Result & Discussion Conclusion & Future Work Hybrid Approaches 18/31 Two Hybrid Approaches TS-GA:: best solution from TS to be considered as an intial solution in GA Require: IntialSolution & TasksPosition & RobotsPosition Solutionbest ← Tabu-Search(IntialSolution,TasksPosition,RobotsPosition) RUN Genetic-Algorithm(Solutionbest ,TasksPosition,RobotsPosition) return Best achieved solution GA-TS:: best solution from GA to be considered as an intial solution in TS Require: IntialSolution & TasksPosition & RobotsPosition Solutionbest ← Genetic-Algorithm(IntialSolution,TasksPosition,RobotsPosition) RUN Tabu-Search(Solutionbest ,TasksPosition,RobotsPosition) return Best achieved solution Mohamed Gomaa Ghanem German University in Cairo Multi-robot Task Allocation Using Meta-heuristic Optimization
  • 55. Introduction Multi-Robot Systems Task Allocation in MRS Proposed Approach Result & Discussion Conclusion & Future Work Hybrid Approaches 18/31 Two Hybrid Approaches TS-GA:: best solution from TS to be considered as an intial solution in GA Require: IntialSolution & TasksPosition & RobotsPosition Solutionbest ← Tabu-Search(IntialSolution,TasksPosition,RobotsPosition) RUN Genetic-Algorithm(Solutionbest ,TasksPosition,RobotsPosition) return Best achieved solution GA-TS:: best solution from GA to be considered as an intial solution in TS Require: IntialSolution & TasksPosition & RobotsPosition Solutionbest ← Genetic-Algorithm(IntialSolution,TasksPosition,RobotsPosition) RUN Tabu-Search(Solutionbest ,TasksPosition,RobotsPosition) return Best achieved solution Mohamed Gomaa Ghanem German University in Cairo Multi-robot Task Allocation Using Meta-heuristic Optimization
  • 56. Introduction Multi-Robot Systems Task Allocation in MRS Proposed Approach Result & Discussion Conclusion & Future Work Hybrid Approaches 18/31 Two Hybrid Approaches TS-GA:: best solution from TS to be considered as an intial solution in GA Require: IntialSolution & TasksPosition & RobotsPosition Solutionbest ← Tabu-Search(IntialSolution,TasksPosition,RobotsPosition) RUN Genetic-Algorithm(Solutionbest ,TasksPosition,RobotsPosition) return Best achieved solution GA-TS:: best solution from GA to be considered as an intial solution in TS Require: IntialSolution & TasksPosition & RobotsPosition Solutionbest ← Genetic-Algorithm(IntialSolution,TasksPosition,RobotsPosition) RUN Tabu-Search(Solutionbest ,TasksPosition,RobotsPosition) return Best achieved solution Mohamed Gomaa Ghanem German University in Cairo Multi-robot Task Allocation Using Meta-heuristic Optimization
  • 57. Introduction Multi-Robot Systems Task Allocation in MRS Proposed Approach Result & Discussion Conclusion & Future Work Hybrid Approaches 18/31 Two Hybrid Approaches TS-GA:: best solution from TS to be considered as an intial solution in GA Require: IntialSolution & TasksPosition & RobotsPosition Solutionbest ← Tabu-Search(IntialSolution,TasksPosition,RobotsPosition) RUN Genetic-Algorithm(Solutionbest ,TasksPosition,RobotsPosition) return Best achieved solution GA-TS:: best solution from GA to be considered as an intial solution in TS Require: IntialSolution & TasksPosition & RobotsPosition Solutionbest ← Genetic-Algorithm(IntialSolution,TasksPosition,RobotsPosition) RUN Tabu-Search(Solutionbest ,TasksPosition,RobotsPosition) return Best achieved solution Mohamed Gomaa Ghanem German University in Cairo Multi-robot Task Allocation Using Meta-heuristic Optimization
  • 58. Introduction Multi-Robot Systems Task Allocation in MRS Proposed Approach Result & Discussion Conclusion & Future Work Hybrid Approaches 18/31 Two Hybrid Approaches TS-GA:: best solution from TS to be considered as an intial solution in GA Require: IntialSolution & TasksPosition & RobotsPosition Solutionbest ← Tabu-Search(IntialSolution,TasksPosition,RobotsPosition) RUN Genetic-Algorithm(Solutionbest ,TasksPosition,RobotsPosition) return Best achieved solution GA-TS:: best solution from GA to be considered as an intial solution in TS Require: IntialSolution & TasksPosition & RobotsPosition Solutionbest ← Genetic-Algorithm(IntialSolution,TasksPosition,RobotsPosition) RUN Tabu-Search(Solutionbest ,TasksPosition,RobotsPosition) return Best achieved solution Mohamed Gomaa Ghanem German University in Cairo Multi-robot Task Allocation Using Meta-heuristic Optimization
  • 59. Introduction Multi-Robot Systems Task Allocation in MRS Proposed Approach Result & Discussion Conclusion & Future Work Hybrid Approaches 18/31 Two Hybrid Approaches TS-GA:: best solution from TS to be considered as an intial solution in GA Require: IntialSolution & TasksPosition & RobotsPosition Solutionbest ← Tabu-Search(IntialSolution,TasksPosition,RobotsPosition) RUN Genetic-Algorithm(Solutionbest ,TasksPosition,RobotsPosition) return Best achieved solution GA-TS:: best solution from GA to be considered as an intial solution in TS Require: IntialSolution & TasksPosition & RobotsPosition Solutionbest ← Genetic-Algorithm(IntialSolution,TasksPosition,RobotsPosition) RUN Tabu-Search(Solutionbest ,TasksPosition,RobotsPosition) return Best achieved solution Mohamed Gomaa Ghanem German University in Cairo Multi-robot Task Allocation Using Meta-heuristic Optimization
  • 60. Introduction Multi-Robot Systems Task Allocation in MRS Proposed Approach Result & Discussion Conclusion & Future Work Hybrid Approaches 18/31 Two Hybrid Approaches TS-GA:: best solution from TS to be considered as an intial solution in GA Require: IntialSolution & TasksPosition & RobotsPosition Solutionbest ← Tabu-Search(IntialSolution,TasksPosition,RobotsPosition) RUN Genetic-Algorithm(Solutionbest ,TasksPosition,RobotsPosition) return Best achieved solution GA-TS:: best solution from GA to be considered as an intial solution in TS Require: IntialSolution & TasksPosition & RobotsPosition Solutionbest ← Genetic-Algorithm(IntialSolution,TasksPosition,RobotsPosition) RUN Tabu-Search(Solutionbest ,TasksPosition,RobotsPosition) return Best achieved solution Mohamed Gomaa Ghanem German University in Cairo Multi-robot Task Allocation Using Meta-heuristic Optimization
  • 61. Introduction Multi-Robot Systems Task Allocation in MRS Proposed Approach Result & Discussion Conclusion & Future Work Hybrid Approaches 18/31 Two Hybrid Approaches TS-GA:: best solution from TS to be considered as an intial solution in GA Require: IntialSolution & TasksPosition & RobotsPosition Solutionbest ← Tabu-Search(IntialSolution,TasksPosition,RobotsPosition) RUN Genetic-Algorithm(Solutionbest ,TasksPosition,RobotsPosition) return Best achieved solution GA-TS:: best solution from GA to be considered as an intial solution in TS Require: IntialSolution & TasksPosition & RobotsPosition Solutionbest ← Genetic-Algorithm(IntialSolution,TasksPosition,RobotsPosition) RUN Tabu-Search(Solutionbest ,TasksPosition,RobotsPosition) return Best achieved solution Mohamed Gomaa Ghanem German University in Cairo Multi-robot Task Allocation Using Meta-heuristic Optimization
  • 62. Introduction Multi-Robot Systems Task Allocation in MRS Proposed Approach Result & Discussion Conclusion & Future Work Experemental Setup 19/31 In all of experiments,some component were used: Java4MRS Simbad-3D Simulator OpenTS & Jgap Matlab The result is an output of running the above programs on a machine with: 32-bit windows Operating System AMD Turion 64 X2 Mobile Technology TL-68 2.4 GHz Processor 3.00 GB of RAM Mohamed Gomaa Ghanem German University in Cairo Multi-robot Task Allocation Using Meta-heuristic Optimization
  • 63. Introduction Multi-Robot Systems Task Allocation in MRS Proposed Approach Result & Discussion Conclusion & Future Work Experemental Setup 19/31 In all of experiments,some component were used: Java4MRS Simbad-3D Simulator OpenTS & Jgap Matlab The result is an output of running the above programs on a machine with: 32-bit windows Operating System AMD Turion 64 X2 Mobile Technology TL-68 2.4 GHz Processor 3.00 GB of RAM Mohamed Gomaa Ghanem German University in Cairo Multi-robot Task Allocation Using Meta-heuristic Optimization
  • 64. Introduction Multi-Robot Systems Task Allocation in MRS Proposed Approach Result & Discussion Conclusion & Future Work Experemental Setup 19/31 In all of experiments,some component were used: Java4MRS Simbad-3D Simulator OpenTS & Jgap Matlab The result is an output of running the above programs on a machine with: 32-bit windows Operating System AMD Turion 64 X2 Mobile Technology TL-68 2.4 GHz Processor 3.00 GB of RAM Mohamed Gomaa Ghanem German University in Cairo Multi-robot Task Allocation Using Meta-heuristic Optimization
  • 65. Introduction Multi-Robot Systems Task Allocation in MRS Proposed Approach Result & Discussion Conclusion & Future Work Experemental Setup 19/31 In all of experiments,some component were used: Java4MRS Simbad-3D Simulator OpenTS & Jgap Matlab The result is an output of running the above programs on a machine with: 32-bit windows Operating System AMD Turion 64 X2 Mobile Technology TL-68 2.4 GHz Processor 3.00 GB of RAM Mohamed Gomaa Ghanem German University in Cairo Multi-robot Task Allocation Using Meta-heuristic Optimization
  • 66. Introduction Multi-Robot Systems Task Allocation in MRS Proposed Approach Result & Discussion Conclusion & Future Work Experemental Setup 19/31 In all of experiments,some component were used: Java4MRS Simbad-3D Simulator OpenTS & Jgap Matlab The result is an output of running the above programs on a machine with: 32-bit windows Operating System AMD Turion 64 X2 Mobile Technology TL-68 2.4 GHz Processor 3.00 GB of RAM Mohamed Gomaa Ghanem German University in Cairo Multi-robot Task Allocation Using Meta-heuristic Optimization
  • 67. Introduction Multi-Robot Systems Task Allocation in MRS Proposed Approach Result & Discussion Conclusion & Future Work Experemental Setup 19/31 In all of experiments,some component were used: Java4MRS Simbad-3D Simulator OpenTS & Jgap Matlab The result is an output of running the above programs on a machine with: 32-bit windows Operating System AMD Turion 64 X2 Mobile Technology TL-68 2.4 GHz Processor 3.00 GB of RAM Mohamed Gomaa Ghanem German University in Cairo Multi-robot Task Allocation Using Meta-heuristic Optimization
  • 68. Introduction Multi-Robot Systems Task Allocation in MRS Proposed Approach Result & Discussion Conclusion & Future Work Experemental Setup 19/31 In all of experiments,some component were used: Java4MRS Simbad-3D Simulator OpenTS & Jgap Matlab The result is an output of running the above programs on a machine with: 32-bit windows Operating System AMD Turion 64 X2 Mobile Technology TL-68 2.4 GHz Processor 3.00 GB of RAM Mohamed Gomaa Ghanem German University in Cairo Multi-robot Task Allocation Using Meta-heuristic Optimization
  • 69. Introduction Multi-Robot Systems Task Allocation in MRS Proposed Approach Result & Discussion Conclusion & Future Work Experemental Setup 19/31 In all of experiments,some component were used: Java4MRS Simbad-3D Simulator OpenTS & Jgap Matlab The result is an output of running the above programs on a machine with: 32-bit windows Operating System AMD Turion 64 X2 Mobile Technology TL-68 2.4 GHz Processor 3.00 GB of RAM Mohamed Gomaa Ghanem German University in Cairo Multi-robot Task Allocation Using Meta-heuristic Optimization
  • 70. Introduction Multi-Robot Systems Task Allocation in MRS Proposed Approach Result & Discussion Conclusion & Future Work Experemental Setup 19/31 In all of experiments,some component were used: Java4MRS Simbad-3D Simulator OpenTS & Jgap Matlab The result is an output of running the above programs on a machine with: 32-bit windows Operating System AMD Turion 64 X2 Mobile Technology TL-68 2.4 GHz Processor 3.00 GB of RAM Mohamed Gomaa Ghanem German University in Cairo Multi-robot Task Allocation Using Meta-heuristic Optimization
  • 71. Introduction Multi-Robot Systems Task Allocation in MRS Proposed Approach Result & Discussion Conclusion & Future Work Evaluation Metrics 20/31 Cost of traveled distance Cost = numberOfRobots i=0 [dist(Ri , T0) + numOfTasks−1forRi j=0 dist(Tj, Tj+1)] Mission completion time = Algorithm Running time + Time taken by last robot to finish its tasks Reliability Scalability Each will be evaluated according to some variables: Number of robots Number of tasks Map area size Algorithm’s parameters Mohamed Gomaa Ghanem German University in Cairo Multi-robot Task Allocation Using Meta-heuristic Optimization
  • 72. Introduction Multi-Robot Systems Task Allocation in MRS Proposed Approach Result & Discussion Conclusion & Future Work Evaluation Metrics 20/31 Cost of traveled distance Cost = numberOfRobots i=0 [dist(Ri , T0) + numOfTasks−1forRi j=0 dist(Tj, Tj+1)] Mission completion time = Algorithm Running time + Time taken by last robot to finish its tasks Reliability Scalability Each will be evaluated according to some variables: Number of robots Number of tasks Map area size Algorithm’s parameters Mohamed Gomaa Ghanem German University in Cairo Multi-robot Task Allocation Using Meta-heuristic Optimization
  • 73. Introduction Multi-Robot Systems Task Allocation in MRS Proposed Approach Result & Discussion Conclusion & Future Work Evaluation Metrics 20/31 Cost of traveled distance Cost = numberOfRobots i=0 [dist(Ri , T0) + numOfTasks−1forRi j=0 dist(Tj, Tj+1)] Mission completion time = Algorithm Running time + Time taken by last robot to finish its tasks Reliability Scalability Each will be evaluated according to some variables: Number of robots Number of tasks Map area size Algorithm’s parameters Mohamed Gomaa Ghanem German University in Cairo Multi-robot Task Allocation Using Meta-heuristic Optimization
  • 74. Introduction Multi-Robot Systems Task Allocation in MRS Proposed Approach Result & Discussion Conclusion & Future Work Evaluation Metrics 20/31 Cost of traveled distance Cost = numberOfRobots i=0 [dist(Ri , T0) + numOfTasks−1forRi j=0 dist(Tj, Tj+1)] Mission completion time = Algorithm Running time + Time taken by last robot to finish its tasks Reliability Scalability Each will be evaluated according to some variables: Number of robots Number of tasks Map area size Algorithm’s parameters Mohamed Gomaa Ghanem German University in Cairo Multi-robot Task Allocation Using Meta-heuristic Optimization
  • 75. Introduction Multi-Robot Systems Task Allocation in MRS Proposed Approach Result & Discussion Conclusion & Future Work Evaluation Metrics 20/31 Cost of traveled distance Cost = numberOfRobots i=0 [dist(Ri , T0) + numOfTasks−1forRi j=0 dist(Tj, Tj+1)] Mission completion time = Algorithm Running time + Time taken by last robot to finish its tasks Reliability Scalability Each will be evaluated according to some variables: Number of robots Number of tasks Map area size Algorithm’s parameters Mohamed Gomaa Ghanem German University in Cairo Multi-robot Task Allocation Using Meta-heuristic Optimization
  • 76. Introduction Multi-Robot Systems Task Allocation in MRS Proposed Approach Result & Discussion Conclusion & Future Work Evaluation Metrics 20/31 Cost of traveled distance Cost = numberOfRobots i=0 [dist(Ri , T0) + numOfTasks−1forRi j=0 dist(Tj, Tj+1)] Mission completion time = Algorithm Running time + Time taken by last robot to finish its tasks Reliability Scalability Each will be evaluated according to some variables: Number of robots Number of tasks Map area size Algorithm’s parameters Mohamed Gomaa Ghanem German University in Cairo Multi-robot Task Allocation Using Meta-heuristic Optimization
  • 77. Introduction Multi-Robot Systems Task Allocation in MRS Proposed Approach Result & Discussion Conclusion & Future Work Evaluation Metrics 20/31 Cost of traveled distance Cost = numberOfRobots i=0 [dist(Ri , T0) + numOfTasks−1forRi j=0 dist(Tj, Tj+1)] Mission completion time = Algorithm Running time + Time taken by last robot to finish its tasks Reliability Scalability Each will be evaluated according to some variables: Number of robots Number of tasks Map area size Algorithm’s parameters Mohamed Gomaa Ghanem German University in Cairo Multi-robot Task Allocation Using Meta-heuristic Optimization
  • 78. Introduction Multi-Robot Systems Task Allocation in MRS Proposed Approach Result & Discussion Conclusion & Future Work Evaluation Metrics 20/31 Cost of traveled distance Cost = numberOfRobots i=0 [dist(Ri , T0) + numOfTasks−1forRi j=0 dist(Tj, Tj+1)] Mission completion time = Algorithm Running time + Time taken by last robot to finish its tasks Reliability Scalability Each will be evaluated according to some variables: Number of robots Number of tasks Map area size Algorithm’s parameters Mohamed Gomaa Ghanem German University in Cairo Multi-robot Task Allocation Using Meta-heuristic Optimization
  • 79. Introduction Multi-Robot Systems Task Allocation in MRS Proposed Approach Result & Discussion Conclusion & Future Work Evaluation Metrics 20/31 Cost of traveled distance Cost = numberOfRobots i=0 [dist(Ri , T0) + numOfTasks−1forRi j=0 dist(Tj, Tj+1)] Mission completion time = Algorithm Running time + Time taken by last robot to finish its tasks Reliability Scalability Each will be evaluated according to some variables: Number of robots Number of tasks Map area size Algorithm’s parameters Mohamed Gomaa Ghanem German University in Cairo Multi-robot Task Allocation Using Meta-heuristic Optimization
  • 80. Introduction Multi-Robot Systems Task Allocation in MRS Proposed Approach Result & Discussion Conclusion & Future Work Evaluation Metrics 20/31 Cost of traveled distance Cost = numberOfRobots i=0 [dist(Ri , T0) + numOfTasks−1forRi j=0 dist(Tj, Tj+1)] Mission completion time = Algorithm Running time + Time taken by last robot to finish its tasks Reliability Scalability Each will be evaluated according to some variables: Number of robots Number of tasks Map area size Algorithm’s parameters Mohamed Gomaa Ghanem German University in Cairo Multi-robot Task Allocation Using Meta-heuristic Optimization
  • 81. Introduction Multi-Robot Systems Task Allocation in MRS Proposed Approach Result & Discussion Conclusion & Future Work Java4MRS Application Mechanism 21/31 Mohamed Gomaa Ghanem German University in Cairo Multi-robot Task Allocation Using Meta-heuristic Optimization
  • 82. Introduction Multi-Robot Systems Task Allocation in MRS Proposed Approach Result & Discussion Conclusion & Future Work Results 22/31 Parameters Tunning Tabu Search Number of iterations Nt ∗ Nr Tabu length 2 √ Nr + Nt Candidates per Iteration max(Nt,Nr ) Genetic Algorithm Number of iterations more is better till (Nr + Nt)2 Population Size more is better till (Nr + Nt)2 Swap Mutation Probability according to the problem(0.9) Mohamed Gomaa Ghanem German University in Cairo Multi-robot Task Allocation Using Meta-heuristic Optimization
  • 83. Introduction Multi-Robot Systems Task Allocation in MRS Proposed Approach Result & Discussion Conclusion & Future Work Results 22/31 Parameters Tunning Tabu Search Number of iterations Nt ∗ Nr Tabu length 2 √ Nr + Nt Candidates per Iteration max(Nt,Nr ) Genetic Algorithm Number of iterations more is better till (Nr + Nt)2 Population Size more is better till (Nr + Nt)2 Swap Mutation Probability according to the problem(0.9) Mohamed Gomaa Ghanem German University in Cairo Multi-robot Task Allocation Using Meta-heuristic Optimization
  • 84. Introduction Multi-Robot Systems Task Allocation in MRS Proposed Approach Result & Discussion Conclusion & Future Work Results 22/31 Parameters Tunning Tabu Search Number of iterations Nt ∗ Nr Tabu length 2 √ Nr + Nt Candidates per Iteration max(Nt,Nr ) Genetic Algorithm Number of iterations more is better till (Nr + Nt)2 Population Size more is better till (Nr + Nt)2 Swap Mutation Probability according to the problem(0.9) Mohamed Gomaa Ghanem German University in Cairo Multi-robot Task Allocation Using Meta-heuristic Optimization
  • 85. Introduction Multi-Robot Systems Task Allocation in MRS Proposed Approach Result & Discussion Conclusion & Future Work Results 22/31 Parameters Tunning Tabu Search Number of iterations Nt ∗ Nr Tabu length 2 √ Nr + Nt Candidates per Iteration max(Nt,Nr ) Genetic Algorithm Number of iterations more is better till (Nr + Nt)2 Population Size more is better till (Nr + Nt)2 Swap Mutation Probability according to the problem(0.9) Mohamed Gomaa Ghanem German University in Cairo Multi-robot Task Allocation Using Meta-heuristic Optimization
  • 86. Introduction Multi-Robot Systems Task Allocation in MRS Proposed Approach Result & Discussion Conclusion & Future Work Results 22/31 Parameters Tunning Tabu Search Number of iterations Nt ∗ Nr Tabu length 2 √ Nr + Nt Candidates per Iteration max(Nt,Nr ) Genetic Algorithm Number of iterations more is better till (Nr + Nt)2 Population Size more is better till (Nr + Nt)2 Swap Mutation Probability according to the problem(0.9) Mohamed Gomaa Ghanem German University in Cairo Multi-robot Task Allocation Using Meta-heuristic Optimization
  • 87. Introduction Multi-Robot Systems Task Allocation in MRS Proposed Approach Result & Discussion Conclusion & Future Work Results 22/31 Parameters Tunning Tabu Search Number of iterations Nt ∗ Nr Tabu length 2 √ Nr + Nt Candidates per Iteration max(Nt,Nr ) Genetic Algorithm Number of iterations more is better till (Nr + Nt)2 Population Size more is better till (Nr + Nt)2 Swap Mutation Probability according to the problem(0.9) Mohamed Gomaa Ghanem German University in Cairo Multi-robot Task Allocation Using Meta-heuristic Optimization
  • 88. Introduction Multi-Robot Systems Task Allocation in MRS Proposed Approach Result & Discussion Conclusion & Future Work Results 22/31 Parameters Tunning Tabu Search Number of iterations Nt ∗ Nr Tabu length 2 √ Nr + Nt Candidates per Iteration max(Nt,Nr ) Genetic Algorithm Number of iterations more is better till (Nr + Nt)2 Population Size more is better till (Nr + Nt)2 Swap Mutation Probability according to the problem(0.9) Mohamed Gomaa Ghanem German University in Cairo Multi-robot Task Allocation Using Meta-heuristic Optimization
  • 89. Introduction Multi-Robot Systems Task Allocation in MRS Proposed Approach Result & Discussion Conclusion & Future Work Results 22/31 Parameters Tunning Tabu Search Number of iterations Nt ∗ Nr Tabu length 2 √ Nr + Nt Candidates per Iteration max(Nt,Nr ) Genetic Algorithm Number of iterations more is better till (Nr + Nt)2 Population Size more is better till (Nr + Nt)2 Swap Mutation Probability according to the problem(0.9) Mohamed Gomaa Ghanem German University in Cairo Multi-robot Task Allocation Using Meta-heuristic Optimization
  • 90. Introduction Multi-Robot Systems Task Allocation in MRS Proposed Approach Result & Discussion Conclusion & Future Work Results 22/31 Parameters Tunning Tabu Search Number of iterations Nt ∗ Nr Tabu length 2 √ Nr + Nt Candidates per Iteration max(Nt,Nr ) Genetic Algorithm Number of iterations more is better till (Nr + Nt)2 Population Size more is better till (Nr + Nt)2 Swap Mutation Probability according to the problem(0.9) Mohamed Gomaa Ghanem German University in Cairo Multi-robot Task Allocation Using Meta-heuristic Optimization
  • 91. Introduction Multi-Robot Systems Task Allocation in MRS Proposed Approach Result & Discussion Conclusion & Future Work Results 22/31 Parameters Tunning Tabu Search Number of iterations Nt ∗ Nr Tabu length 2 √ Nr + Nt Candidates per Iteration max(Nt,Nr ) Genetic Algorithm Number of iterations more is better till (Nr + Nt)2 Population Size more is better till (Nr + Nt)2 Swap Mutation Probability according to the problem(0.9) Mohamed Gomaa Ghanem German University in Cairo Multi-robot Task Allocation Using Meta-heuristic Optimization
  • 92. Introduction Multi-Robot Systems Task Allocation in MRS Proposed Approach Result & Discussion Conclusion & Future Work Results 22/31 Parameters Tunning Tabu Search Number of iterations Nt ∗ Nr Tabu length 2 √ Nr + Nt Candidates per Iteration max(Nt,Nr ) Genetic Algorithm Number of iterations more is better till (Nr + Nt)2 Population Size more is better till (Nr + Nt)2 Swap Mutation Probability according to the problem(0.9) Mohamed Gomaa Ghanem German University in Cairo Multi-robot Task Allocation Using Meta-heuristic Optimization
  • 93. Introduction Multi-Robot Systems Task Allocation in MRS Proposed Approach Result & Discussion Conclusion & Future Work Results 22/31 Parameters Tunning Tabu Search Number of iterations Nt ∗ Nr Tabu length 2 √ Nr + Nt Candidates per Iteration max(Nt,Nr ) Genetic Algorithm Number of iterations more is better till (Nr + Nt)2 Population Size more is better till (Nr + Nt)2 Swap Mutation Probability according to the problem(0.9) Mohamed Gomaa Ghanem German University in Cairo Multi-robot Task Allocation Using Meta-heuristic Optimization
  • 94. Introduction Multi-Robot Systems Task Allocation in MRS Proposed Approach Result & Discussion Conclusion & Future Work Results 22/31 Parameters Tunning Tabu Search Number of iterations Nt ∗ Nr Tabu length 2 √ Nr + Nt Candidates per Iteration max(Nt,Nr ) Genetic Algorithm Number of iterations more is better till (Nr + Nt)2 Population Size more is better till (Nr + Nt)2 Swap Mutation Probability according to the problem(0.9) Mohamed Gomaa Ghanem German University in Cairo Multi-robot Task Allocation Using Meta-heuristic Optimization
  • 95. Introduction Multi-Robot Systems Task Allocation in MRS Proposed Approach Result & Discussion Conclusion & Future Work Results 22/31 Parameters Tunning Tabu Search Number of iterations Nt ∗ Nr Tabu length 2 √ Nr + Nt Candidates per Iteration max(Nt,Nr ) Genetic Algorithm Number of iterations more is better till (Nr + Nt)2 Population Size more is better till (Nr + Nt)2 Swap Mutation Probability according to the problem(0.9) Mohamed Gomaa Ghanem German University in Cairo Multi-robot Task Allocation Using Meta-heuristic Optimization
  • 96. Introduction Multi-Robot Systems Task Allocation in MRS Proposed Approach Result & Discussion Conclusion & Future Work Results 23/31 Number Of Tasks Mohamed Gomaa Ghanem German University in Cairo Multi-robot Task Allocation Using Meta-heuristic Optimization
  • 97. Introduction Multi-Robot Systems Task Allocation in MRS Proposed Approach Result & Discussion Conclusion & Future Work Results 24/31 Number Of Tasks Mohamed Gomaa Ghanem German University in Cairo Multi-robot Task Allocation Using Meta-heuristic Optimization
  • 98. Introduction Multi-Robot Systems Task Allocation in MRS Proposed Approach Result & Discussion Conclusion & Future Work Results 25/31 Number Of Tasks Mohamed Gomaa Ghanem German University in Cairo Multi-robot Task Allocation Using Meta-heuristic Optimization
  • 99. Introduction Multi-Robot Systems Task Allocation in MRS Proposed Approach Result & Discussion Conclusion & Future Work Results 26/31 Number Of Robots Mohamed Gomaa Ghanem German University in Cairo Multi-robot Task Allocation Using Meta-heuristic Optimization
  • 100. Introduction Multi-Robot Systems Task Allocation in MRS Proposed Approach Result & Discussion Conclusion & Future Work Results 27/31 Number Of Robots Mohamed Gomaa Ghanem German University in Cairo Multi-robot Task Allocation Using Meta-heuristic Optimization
  • 101. Introduction Multi-Robot Systems Task Allocation in MRS Proposed Approach Result & Discussion Conclusion & Future Work Results 28/31 Number Of Robots Mohamed Gomaa Ghanem German University in Cairo Multi-robot Task Allocation Using Meta-heuristic Optimization
  • 102. Introduction Multi-Robot Systems Task Allocation in MRS Proposed Approach Result & Discussion Conclusion & Future Work Conclusion 29/31 Experimental results showed that it’s better to use the Genetic Algorithm if an initial solution is not known, but if there exists an initial solution and trying to reach a better one therefor the tabu search outperforms genetic algorithm as it tries to improve the allocation in each iteration and to escape from a local minimum to a global minimum Tabu Search is also better in solving for real time as it provides faster response A hybrid approachs are examined and found that it’s better to get the initial solution from Genetic Algorithm and use it to start Tabu search in order to find out if there exist a better near solutions Mohamed Gomaa Ghanem German University in Cairo Multi-robot Task Allocation Using Meta-heuristic Optimization
  • 103. Introduction Multi-Robot Systems Task Allocation in MRS Proposed Approach Result & Discussion Conclusion & Future Work Conclusion 29/31 Experimental results showed that it’s better to use the Genetic Algorithm if an initial solution is not known, but if there exists an initial solution and trying to reach a better one therefor the tabu search outperforms genetic algorithm as it tries to improve the allocation in each iteration and to escape from a local minimum to a global minimum Tabu Search is also better in solving for real time as it provides faster response A hybrid approachs are examined and found that it’s better to get the initial solution from Genetic Algorithm and use it to start Tabu search in order to find out if there exist a better near solutions Mohamed Gomaa Ghanem German University in Cairo Multi-robot Task Allocation Using Meta-heuristic Optimization
  • 104. Introduction Multi-Robot Systems Task Allocation in MRS Proposed Approach Result & Discussion Conclusion & Future Work Conclusion 29/31 Experimental results showed that it’s better to use the Genetic Algorithm if an initial solution is not known, but if there exists an initial solution and trying to reach a better one therefor the tabu search outperforms genetic algorithm as it tries to improve the allocation in each iteration and to escape from a local minimum to a global minimum Tabu Search is also better in solving for real time as it provides faster response A hybrid approachs are examined and found that it’s better to get the initial solution from Genetic Algorithm and use it to start Tabu search in order to find out if there exist a better near solutions Mohamed Gomaa Ghanem German University in Cairo Multi-robot Task Allocation Using Meta-heuristic Optimization
  • 105. Introduction Multi-Robot Systems Task Allocation in MRS Proposed Approach Result & Discussion Conclusion & Future Work Conclusion 29/31 Experimental results showed that it’s better to use the Genetic Algorithm if an initial solution is not known, but if there exists an initial solution and trying to reach a better one therefor the tabu search outperforms genetic algorithm as it tries to improve the allocation in each iteration and to escape from a local minimum to a global minimum Tabu Search is also better in solving for real time as it provides faster response A hybrid approachs are examined and found that it’s better to get the initial solution from Genetic Algorithm and use it to start Tabu search in order to find out if there exist a better near solutions Mohamed Gomaa Ghanem German University in Cairo Multi-robot Task Allocation Using Meta-heuristic Optimization
  • 106. Introduction Multi-Robot Systems Task Allocation in MRS Proposed Approach Result & Discussion Conclusion & Future Work Future Work 30/31 Robots and Tasks heterogeneity More Robot Planning (obstical avoidance) Integrate Google Map Input ROS(Robot Operating System) insted of Simbad-3D Mohamed Gomaa Ghanem German University in Cairo Multi-robot Task Allocation Using Meta-heuristic Optimization
  • 107. Introduction Multi-Robot Systems Task Allocation in MRS Proposed Approach Result & Discussion Conclusion & Future Work Future Work 30/31 Robots and Tasks heterogeneity More Robot Planning (obstical avoidance) Integrate Google Map Input ROS(Robot Operating System) insted of Simbad-3D Mohamed Gomaa Ghanem German University in Cairo Multi-robot Task Allocation Using Meta-heuristic Optimization
  • 108. Introduction Multi-Robot Systems Task Allocation in MRS Proposed Approach Result & Discussion Conclusion & Future Work Future Work 30/31 Robots and Tasks heterogeneity More Robot Planning (obstical avoidance) Integrate Google Map Input ROS(Robot Operating System) insted of Simbad-3D Mohamed Gomaa Ghanem German University in Cairo Multi-robot Task Allocation Using Meta-heuristic Optimization
  • 109. Introduction Multi-Robot Systems Task Allocation in MRS Proposed Approach Result & Discussion Conclusion & Future Work Future Work 30/31 Robots and Tasks heterogeneity More Robot Planning (obstical avoidance) Integrate Google Map Input ROS(Robot Operating System) insted of Simbad-3D Mohamed Gomaa Ghanem German University in Cairo Multi-robot Task Allocation Using Meta-heuristic Optimization
  • 110. Introduction Multi-Robot Systems Task Allocation in MRS Proposed Approach Result & Discussion Conclusion & Future Work Future Work 30/31 Robots and Tasks heterogeneity More Robot Planning (obstical avoidance) Integrate Google Map Input ROS(Robot Operating System) insted of Simbad-3D Mohamed Gomaa Ghanem German University in Cairo Multi-robot Task Allocation Using Meta-heuristic Optimization
  • 111. Introduction Multi-Robot Systems Task Allocation in MRS Proposed Approach Result & Discussion Conclusion & Future Work 1. download from : https://javar.googlecode.com/files/Java4Robot.jar 2. run & enjoy ... Demo Mohamed Gomaa Ghanem German University in Cairo Multi-robot Task Allocation Using Meta-heuristic Optimization