![Signal Design for an Isolated Intersection during Congestion
Talmor, I
Mahalel, D
In this paper, the two authors present another signal design for isolated, oversaturated
intersections. They aim to “alleviate long queues during severe congestion conditions, which
cause both lengthy delays and harsh environmental damage.” To solve this, they ask the
question: what is the best way to relieve queues during oversaturated conditions at isolated
intersections?
To this, the authors answer: by maximizing the average throughput of the intersection,
the “number of vehicles in the queue is reduced at the fastest possible rate and the period of
congestion is shortened.” In this paper, the author starts off by referencing past research on the
same topic and points out that many past research prevents adequate signal plans from being
constructed as they assume that saturation flow (“vehicle’s discharging rate from an existing
queue”) is constant. This is true, as in real intersections, the saturation flow decreases as the
green time lengthens, as cars need time to accelerate and therefore do not have equal gaps
between them. The paper explains that saturation flow may have been kept constant in past
research, as the “green light (be) terminated before the decline in the flow [began].” The authors
also point out that most past studies aim to minimize delays, which require “advance
knowledge of demands, queue lengths etc.,” but that “this information is often lacking during
unexpected severe conditions of oversaturation.” They also claim that the objective of
minimizing delay “might not be appropriate when the demand is greater than the average
capacity, because the queues lengthen and, consequently, the delays grow rapidly.” The paper
also indicates that “it is better to maximize cycle throughput rather than minimize delay,” as
“the more vehicles that cross the intersection per time unit, the less the total waiting time is,”
and therefore “minimizing total delay can be achieved simply by maximizing total throughput.”
Therefore, the authors propose an algorithm which aims to maximize average and per-cycle
throughput (actual number of vehicles per time unit that can cross the intersection, given the
signal plan) “by balancing two opposing aspects—lost time and declining saturation flow—
without neglecting essential constraints.” (lost time is when no cars can go through the
intersection) Through this algorithm, the researchers calculated the cycle time for each
approach which maximizes the throughput of the intersection which decreased the congestion
period and delay. They also found out that “saving lost time is better achieved when the green
lights are extended to the range in which the applied discharge flows decrease.”
This research yet provides another new method to approach solving congestions at
intersections. This time, the paper focuses on maximizing the number of cars that can pass
through the intersection rather than minimizing delay times or queue lengths. This allows the
method to function without many variables needed in other factors such as queue length or
demand of intersection usage. This makes this method more reliable, exact, and cheap
compared to the previous methods discussed, but one drawback is that queue lengths have to
be sufficiently long enough that the whole queue cannot be resolved in one cycle for the
algorithm to function.](https://image.slidesharecdn.com/6b09f1e3-a190-49b7-9593-1787dfa5d0a9-160921012151/85/Research-6-320.jpg)
Research
- 1. Identification of Oversaturated Intersections Using High-Resolution Traffic Signal Data Xinkai Wu Henry X. Liu Douglas Gettman In this article, the authors aim to explore an objective method of measuring congestion at intersections. The authors clarify this by writing, “This research takes one step further toward a better understanding of oversaturation, by providing a coherent methodology to detect the onset of oversaturation at signalized intersections.” From this, we can get the authors’ overall question: what is a practical and effective way to measure and detect oversaturation at signalized intersections? To this question, the authors answer: quantifying the detrimental effects of oversaturation gives a practical way to measure and detect oversaturation at signalized intersections. To understand how the authors got to this conclusion, we must first understand the concept of oversaturation, which they define as, traffic demand exceeding the capacity of a facility. The authors point out that it is impossible to apply this definition directly to quantify oversaturation, as the definition is abstract, which is why they propose a circumventing method of measuring the detrimental effects of oversaturation. The authors define detrimental effects of oversaturation in the temporal dimension, which is represented by a residual queue at the end of a signal cycle. From this, the authors define the oversaturation severity index (OSI): “the ratio between unusable green time and total available green time in a cycle” with a non- negative percentage value between 0-100 with 100 meaning that no green time is usable. The unusable green time is defined as the “green time to discharge the residual queue in the following cycle.” This source helps to address the question regarding the method of evaluating congestion in a quantitative way. By providing a way to quantify a previously qualitative measurement, this research helps to support future research explore instances of congestion and solutions to oversaturation. The source also gives a definition of oversaturation: “Conceptually, an oversaturated traffic intersection is defined as one where traffic demand exceeds the capacity.” This helps to establish the meaning of the core problem that the practical problem gives. Therefore, from here on, we can understand what future research indicates by oversaturation and congestion.
- 2. Optimum Traffic Control-Optimization Criteria Gordana Senborn Bratislav Lazic Slodoban Guberinic In this book, the authors aim to provide various methods to optimize traffic at isolated signalized interactions. To be specific, the book “is concerned with the traffic control problem on a single, isolated signalized intersection.” This clearly gives the authors’ question: how does one optimally control traffic in a single, isolated signalized intersection? To this question, the authors answer: for each intersection, different strategies must be applied, with each strategy determined by different factors and situations. The authors support this thesis by first providing an explanation and parts of an intersection and moves on to different factors and various equations that can be used in different situation. Simply, the authors have compiled a plethora of research and added their opinions on which situation each category should be used. I will be focusing on chapter 9 of this book, Optimization criteria. In this chapter, the authors try to answer the question: what goal needs to be achieved in designing each intersection? They answer this question by asserting that for each intersection, each should have an optimization goal. To start, the authors give an outline to what criterion should be met for each criteria. Each criteria should be measurable, related to the intersection as a whole in a given time, and it should be an explicit function. Some criteria that are used most often in intersection design and improvement include total delay, number of vehicle stops, fuel consumption, and environmental influence. These usually fall into three categories: criteria related to capacity, queuing, and environment. The authors mention “it would be desirable to choose the control that optimizes all of the mentioned criteria, but this is not possible,” which is why different criteria for signal planning are applied in different traffic conditions. In addition, the authors also claim that duration of congestion and queue lengths are other important factors that could be considered in an oversaturated intersection, as these factors take priority compared to other criteria such as the environment in congested situations. This source serves to answer the question on what are the variables that are considered during intersection design and signal timing methods. This information helps us to analyze each method through what criteria they optimize and therefore we can understand under what conditions each method should be used.
- 3. Optimal Signal Timing for an Oversaturated Intersection (Minimum Delay) Tang-Hsien Chang Jen-Ting Lin In this article, the authors propose a novel method of determining the signal times in an oversaturated intersection. The authors claim that the paper presents “a timing decision methodology which considers the whole oversaturation period,” which differs from other commonly used methods. This shows the authors’ question: what is a method to improve on previous methods to determine the signal timing of an intersection that covers the whole oversaturation period? The authors answer this question: two methods that cover the whole oversaturation period have been developed, one being the discrete minimal delay model and the other being the performance index model, with the performance index model being a more appropriate design. First, we will examine the discrete minimal delay model. This model serves to “manifest the complexity of the continuous delay model developed by Michalopoulos and Stephanopolos.” The continuous delay model aims to find the optimal switch-over point during the oversaturation period to minimize delay of traffic. The switch-over point is defined as a point whereàInitially, the maximum green time is given to the approach with the greatest arrival rate and the minimum green time given to the other approach with the smaller arrival rate. At the switch-over point, the maximum green time is given to the approach with the smaller arrival rate and the minimum green time is given to the approach with the greatest arrival rate to resolve the queue at the approach with the smaller arrival rate. Here, the authors exerts that “continuous type models are limited in that the switch-over point does not necessarily occur at the end of a cycle,” whereas “switch-over points determined by a discrete model occur exactly at the termination of a cycle.” With this explanation, the authors demonstrates the benefits and differences of the discrete delay model compared to the continuous model and also points out that the discrete model demonstrates “that pure delay models are ineffective in searching optimal cycle length.” Second, we have the performance index model. The authors introduces this model by suggesting that this model is “more appropriate in studying oversaturation control.” The performance index model is simply a variation of the discrete minimal delay model, with the difference of the performance index model being that it incorporates vehicle stop factors. The authors write that because during oversaturation the number of stops and arrivals increase in an intersection, the inclusion of vehicle delays by stopping will make the model more accurate. In the conclusion after the tests, the author claims, “conventional signal control strategies are inadequate since the designed optimization timing is only considered for the next single cycle after the executing one, not concurrently for the entire congestion period,” and that the two methods covered in the paper “significantly outperforms conventional equal time-sharing dispersion control.” This source serves as one strategy to employ at an oversaturated intersection. This source tries to improve congestion at intersections by considering a switch-over point to determine which approach should receive the maximum or minimum green time over the course of the whole congestion period rather than a single cycle. This method is an improvement over conventional methods, as it incorporates the whole oversaturation period as well as vehicle stop factors.
- 4. Reexamining Vehicle-Actuation Strategies at Isolated Signalized Intersections (Minimum Delay) Michael Cassidy Yu-Hao Chuang Jeff Vitale In this paper, the authors aim to “demonstrate potential benefits from deploying efficient vehicle actuation strategies” that have been developed by Kell and Fullerton 1991. They do this through a “simulation to assess intersection performance under enhanced vehicle-actuated (VA) signal control.” Therefore, the question that the author aims to answer would be: what are the effects/benefits from deploying efficient VA strategies at isolated intersections? To this question, the authors answer that experiments indicated that enhanced (efficient) VA strategies reduced vehicle delay in excess of 30% in some situations compared to standard VA strategies. To understand this clearly, we must first examine what a vehicle actuated signalized intersection means. This simply means that there are detectors placed in the intersection that are able to detect the presence and passing of vehicles, which in turn affects the signal controller to determine phase times of each approach. The authors also explain what advantages general VA controls offer compared to fixed-time signalization, which include the “capability to respond to cyclic fluctuations in arrival rate” and “the capability to reduce the lost time incurred when changing signal indications.” From here, the difference between conventional VA methods and enhanced VA methods emerge. The conventional VA method fails to exploit the two benefits fully, as this method relies on a ‘critical gap’ to indicate that the queue has fully discharged. This critical gap is the time where no vehicles pass through the detector at the front of the intersection. This causes unnecessary loss of time and efficiency, as a certain amount of green time must pass to let the signal change back to red. The enhanced VA method aims to fix this issue by placing the detector so that the end of a queue can be detected before it enters the intersection, which allows only the queue to pass through the intersection before the light changes to red immediately. This reduces delay in the intersection and therefore also reduces the overall oversaturation period. This paper helps to provide another potential method to employ at oversaturated intersections. This method shown in this research is a common strategy that many people are knowledgeable about. The use of this source however, is to show that a common method such as this may not be appropriate to certain intersections, especially the one covered in the practical problem. The enhanced VA control may not be suitable in constantly oversaturated intersections, as when the queue does not vanish in one cycle, the “VA control consistently displays maximum cycle times thus providing no advantage over fixed-time control.” In addition, “although VA control can enhance performance at “low flow” intersections, the cost of VA equipment may not warrant deployment at such locations.”
- 5. Optimal Steady-State Control for Isolated Traffic Intersections Jack Haddad Bart De Schutter Ilya Ioslovich Per-Olof Gutman The authors of this paper aim to propose another improved method to optimize an isolated intersection. To do that, they first asked the question: what is the best way to optimize an intersection with more exact results and simper calculations? From the question, they propose “a discrete-event max-plus model…to formulate an optimization problem for the green-red switching sequence.” Before we can understand the model, we must first look at where the authors are approaching the problem. The authors first mention other papers that have tackled the problem of controlling oversaturated isolated intersections such as Optimal Signal Timing for an Oversaturated Intersection and mention that these papers’ aims were to minimize delays. The authors continue to say that now, with the development of sensors, it is easier to measure queue lengths than estimate delays through the formulas given. They also point out that using sensor data results in more accurate results compared to delay estimation techniques previously developed. Therefore, this paper focuses on a method to relieve congestion through minimizing queue lengths on each approach. This method, like previously developed methods, keep total throughput (#of vehicles crossing the intersection), cycle length, saturation flows (maximum flow from a particular approach) and arrival rates constant. Meanwhile, the paper still aims to find the perfect timing to change the signals on the intersection, just like the two previous sources, but uses a different optimization category of queue length. To find the formula, the authors use queue length and minimize it to find the function, which yields optimal signal times for each approach. This source provides another way to alleviate congestion in an oversaturated intersection. This method is different from other methods previously discussed, as this methods optimizes queue length rather than delay times, which was the case for the previous two methods. Therefore, this source proposes an alternative method that is not a direct improvement from the previous two methods and provides a different viewpoint to the problem. A point that should be considered is that this method requires a sensor to be feasible while the other method that was discussed did not require one. Here, there is a tradeoff between money and accuracy/efficiency. In the case of the practical problem we are discussing, we must analyze if the money is worth the increased accuracy.
- 6. Signal Design for an Isolated Intersection during Congestion Talmor, I Mahalel, D In this paper, the two authors present another signal design for isolated, oversaturated intersections. They aim to “alleviate long queues during severe congestion conditions, which cause both lengthy delays and harsh environmental damage.” To solve this, they ask the question: what is the best way to relieve queues during oversaturated conditions at isolated intersections? To this, the authors answer: by maximizing the average throughput of the intersection, the “number of vehicles in the queue is reduced at the fastest possible rate and the period of congestion is shortened.” In this paper, the author starts off by referencing past research on the same topic and points out that many past research prevents adequate signal plans from being constructed as they assume that saturation flow (“vehicle’s discharging rate from an existing queue”) is constant. This is true, as in real intersections, the saturation flow decreases as the green time lengthens, as cars need time to accelerate and therefore do not have equal gaps between them. The paper explains that saturation flow may have been kept constant in past research, as the “green light (be) terminated before the decline in the flow [began].” The authors also point out that most past studies aim to minimize delays, which require “advance knowledge of demands, queue lengths etc.,” but that “this information is often lacking during unexpected severe conditions of oversaturation.” They also claim that the objective of minimizing delay “might not be appropriate when the demand is greater than the average capacity, because the queues lengthen and, consequently, the delays grow rapidly.” The paper also indicates that “it is better to maximize cycle throughput rather than minimize delay,” as “the more vehicles that cross the intersection per time unit, the less the total waiting time is,” and therefore “minimizing total delay can be achieved simply by maximizing total throughput.” Therefore, the authors propose an algorithm which aims to maximize average and per-cycle throughput (actual number of vehicles per time unit that can cross the intersection, given the signal plan) “by balancing two opposing aspects—lost time and declining saturation flow— without neglecting essential constraints.” (lost time is when no cars can go through the intersection) Through this algorithm, the researchers calculated the cycle time for each approach which maximizes the throughput of the intersection which decreased the congestion period and delay. They also found out that “saving lost time is better achieved when the green lights are extended to the range in which the applied discharge flows decrease.” This research yet provides another new method to approach solving congestions at intersections. This time, the paper focuses on maximizing the number of cars that can pass through the intersection rather than minimizing delay times or queue lengths. This allows the method to function without many variables needed in other factors such as queue length or demand of intersection usage. This makes this method more reliable, exact, and cheap compared to the previous methods discussed, but one drawback is that queue lengths have to be sufficiently long enough that the whole queue cannot be resolved in one cycle for the algorithm to function.
- 7. Review of Traffic Signal Control Methods Under Over-Saturated Conditions Li Yan Zhao Zhi-hong Li Peng-fei Huang Shen-sen Chen Kuan-min In this research, the authors aim to provide a comprehensive review and commentary on developed traffic signal control methods. They ask the question: what categories do each traffic signal control method fall under, what function do they serve, how to they function, when are they efficient/inefficient, and what method should one use in each case? They respond to this question: “to provide references for traffic control optimization at congested urban road network, various traffic signal control strategies under over-saturated conditions were reviewed and evaluated. Based on the analysis of the characteristics of over- saturated traffic flow, the corresponding optimization principles of traffic signal control under over-saturation conditions were proposed.” To this, the authors continue explaining, “the traffic control strategies for isolated intersection, coordinated intersections and road network were summarized and evaluated by using VISSIM simulation environment.” “Simulation result indicates that traffic signal control strategies under over-saturated conditions should optimize the allocation of road space in the first place. The queue management strategy has better performance on optimizing traffic signal under over-saturated conditions.” From this paper, I will use section 2 Over-saturated control strategies for isolated intersection. Here, the authors ask: what is the most fitting strategy in each oversaturated intersection? To this question, the authors answer: the best strategy for over-saturated intersections that are congested in all sides are strategies that optimize queue length or maximize throughput of intersections. To gain insight to how the authors arrived to this conclusion, we must first see the categories that the authors gave. The first category that the authors give is the fixed signal control strategy. They claim that this method “can hardly be utilized under over-saturated conditions,” as this method can do nothing except simply allocate more green time. The second category includes methods that minimize total delays. Examples of these include VA signal control, and the first method covered in this paper. The authors call this the switching of green strategy and realizes that it improves the efficiency of intersections especially when only one or two movements are congested. “However, when the volumes of all movements exceed their capacity, these algorithms will extend all phases to maximum green time” and “at that time, traffic control system has no differences with fixed time control.” The third method is called the ‘maintaining queue ratio’ strategy and incudes methods that optimize queue length, minimize detrimental effects (from the first paper), and maximize throughput. The authors recommend this method as they are generally more accurate, responsive, and universally applicable in that they require less variables and use sensor data rather than estimations. They back their recommendation by providing simulation results in a four-way intersection with varying traffic volumes at each approach. The authors write, that under normal conditions, VA control methods, delay minimizing methods, and maintaining queue ratio strategies faired about the same with fixed timing control strategies lagging behind. “However, under over- saturated conditions, the queue ratio maintenance method has better performance than all other strategies, and thus, it is suitable to be utilized under congested conditions.”
- 8. This research serves my research as a general review for all the methods that have been covered in this research. While the individual methods had to be covered to explore details, this article was covered to gain a general insight about each strategy and to look from another perspective at various strategies that have been covered. This research also provides independent simulation results, which strengthens the authors’ claims. Overall to summarize, the authors claim that queue ratio maintenance usually works best, but delay minimizing methods could be a cheap alternative for intersections that aren’t fully congested on all sides.