Comparative Analysis of Power Quality Enhancement of Distribution System using Unified Power Quality Conditioner

International Research Journal of Engineering and Technology (IRJET) e-ISSN: 2395-0056
Volume: 10 Issue: 03 | Mar 2023 www.irjet.net p-ISSN: 2395-0072
© 2023, IRJET | Impact Factor value: 8.226 | ISO 9001:2008 Certified Journal | Page 646
Comparative Analysis of Power Quality Enhancement of Distribution
System using Unified Power Quality Conditioner
Sobhanjit Padhi1, Rudra Prasad Swain2, Arishmita Banerjee 3 Abhimanyu Mohapatra 4
1, 2, 3 U.G. Student, Department of Electrical Engineering, Odisha University of Technology and Research
4Assistant Professor, Department of Electrical Engineering, Odisha University of Technology and Research
---------------------------------------------------------------------***---------------------------------------------------------------------
Abstract - The increased use of nonlinearelectronicloads is
the primary driver behind the focus on the quality of electric
power. The utilities are finding it difficult to deal with the
different power quality challenges such as voltage swell, sag,
interruptions, harmonic current, andvoltageharmonics. Many
issues are brought on by current harmonics, including an
increase in power system losses, conductor overheating, a
burden of reactive power, relay malfunctions, and low power
factor. The burden of ensuring power quality is therefore
constantly enormous. Electric companies and end users are
now concentrating on the quality of electricity. The economic
benefit is the main driver behind concentrating on electrical
power quality. The use of custom power devices is taken into
account to protect such delicate loads. The quality and
dependability of the power given to clients are improved
through custom power. Low harmonic distortion, a reduction
in supply voltage flicker, and a decrease in voltage sag and
swell are all effects of improved power quality. Thus, in this
paper, Unified Power Quality Conditioner, widely known as
UPQC is modeled and analyzed through various control
strategies to mitigate the power quality issues of a 16-bus
distribution System and the results have been compared
accordingly.
Key Words: Power Quality, Voltage Sag, Harmonics,
UPQC, PI Controller, Neural Network, ANFIS
1. INTRODUCTION
The power system is constantly evolving and poses
various challenges for electrical engineers, particularly in
terms of power quality. With the increasing use of sensitive
equipment and non-linear loads, voltage sag and harmonics
have emerged as the most significant power quality issues,
which are influenced by loads and switching circuits. In
response to these challenges, active power filters have been
developed and widely adopted as means to compensate for
current- and voltage-based distortions. The Unified Power
Quality Conditioner (UPQC) is a particularly advanced
solution that combines both shunt and series APFs to
mitigate power quality issues. By addressing both current
and voltage distortions, UPQC offers a comprehensive
solution for end-users seeking to improvepower qualityand
ensures the stable operation of their electrical systems.
The main objective of this paper is to analyze the Unified
Power Quality Conditioner (UPQC) for improving power
quality in the 16-bus distribution system proposed by S.
Civanlar, using different types of controllers such as PI,
Neural Network, and ANFIS. The modeling of the UPQC has
been carried out and the outputs obtained from the various
controllers have been compared. [1]
2.UNIFIEDPOWERQUALITYCONDITIONER(UPQC)
Thepowersystemhasbeenimproved byintegrating
the Unified Power Quality Conditioner (UPQC), which
provides fast and efficient compensation of both active and
reactive power. As a hybrid Active Power Filter (APF),UPQC
is a versatile device capable of mitigating multiple power
quality issues related to both voltage and current
simultaneously
UPQC = Series APF + Shunt APF
The Unified Power Quality Conditioner (UPQC) is a
combination of shunt and series Active Power Filters(APFs)
that can simultaneously mitigate voltage distortions on the
supply side and current harmonics on the load side. As a
result, it can provide pure sinusoidal voltage and current at
both the supply and load ends
Fig.1: Block Diagram of UPQC
Apart from the series APF and shunt APF, the other
components of UPQC are: -
 DC link: The charge-storing capacitorcanbeusedas
a common DC link that supplies the DC voltage.
 LC filter: The high switching ripples generated by
the series active filter are minimized by the low-

pass LC filter. The LC filter acts as the low-pass
filter.
 Lsh Filter: The ripples generated during switching
mode are minimized by the high-pass Lsh filter.
 Injection transformer: The series injection
transformer is used for voltage injection to the
power system.
2.1 Series Control Strategy
In a transmission line, series APF is generally
connected in series. The series Active Power Filter is a
voltage source inverter connected in series with the
transmission line through a transformer named as an
injection transformer. The series control strategy is a
method used to address and mitigate problems caused by
voltage distortions and imbalances. It involves injecting a
compensating voltage to regulate and balance the load
voltage. The series control strategy involves comparing the
source voltage (VS) and load voltage (VL) to determine the
necessary feedback voltage. A PLL is used to maintain the
system frequency consistency. The feedback voltage's
magnitude and phase are then determined, and a reference
voltage (Va, Vb, Vc) is calculated. The compensating voltage
is then injected back into the system using an insertion
transformer. By the use of PWM (Pulse Width Modulation)
technology, series inverters are controlled. It functions as a
generator of regulated voltage. It is capable of compensating
for voltage imbalances, regulating the voltage at the utility
consumer point of common coupling (PCC).
2.2 Shunt Control Strategy
In transmission lines, shunt Active Power Filters
(APFs) are typically connected in parallel to compensate for
distortions and harmonics caused by the flow of current.
These harmonics are often produced by non-linear loads,
and shunt APFs are used to keep the source current
completely sinusoidal and free from distortion. The shunt
control strategy utilizes both the source voltage (VS) and
load current (IL) to calculate active power (P) and reactive
power (Q). To convert the 3-phase source voltage (VS) into
2-phase values (Vα and Vβ), Clark's transformation
technique is applied.
= ………. (1)
Using PQ theory, the reference current (Iα, Iβ) can
be calculated in αβ coordinates.
= ………. (2)
The load side instantaneous real and imaginary
power componentscanbecalculatedusingloadcurrentsand
phase-neutral voltages as follows:
= ………. (3)
The load side instantaneous real and imaginary powers
comprise both AC and DC components. The DC components
of power (p) and reactive power (q) are made up of the
positive sequence components ( and ) of the load current.
The AC components of active power ( ) and reactive power
( ) comprise harmonic and negative sequence components
of the load currents.
………. (4)
A PI controller compares the reference voltage with the
voltage flowing across the capacitor and outputs the power
lost proportionally to the error produced by comparing the
two.
= ..(5)
Where the reference currents are in αβ coordinates. The
reference current is then transformed back into abc
coordinates using inverse Clark's transformation.
= ………. (6)
The reference current (Ia*, Ib*, Ic*) is compared with the
current flowing through the shunt controller,andtheoutput
is passed through a hysteresis controller to generate gate
pulse signals for the IGBTs within the shunt inverter. This
produces a harmonic-free current.
3. CONTROLLERS USED
3.1 PI Controller
The PI controller is a commonly used feedback
mechanism in various industrial control systems and other
applications.Itemploystwoparameters:theProportional(P)
and the Integral (I) parameters. The P parametergovernsthe
controller's response to the current error, whereas the I
parameter determines the response based on the

accumulation of past errors. This dual approach allows the
controller to provide a stable and accurate response to
dynamic changes in the system. Mathematically, the PI
controller can be expressed as an equation in the following
form:
The output of the PI controller representedasY(t),canbe
calculated using an equation that takes into account the
proportional gain Kp, integral gain Ki, and the input voltage
error value e (t).
The main objective of the PI controller is to maintain the
DC-link voltage at a specifiedreferencevalue.Toachievethis,
the DC-link capacitormust be suppliedwithacertainamount
of real power, which is proportional to the difference
between the actual voltage and the reference voltage. By
controlling this powerflow,thePIcontrollerhelpstoregulate
the DC-link voltage and ensure the stable operation of the
system.
The power that is needed by the DC-link capacitor can be
mathematically expressed as follows:
When the proportional gain (Kp) and integral gain (Ki)
values are set to be large, the regulation of the DC-bus
voltage becomes dominant and the steady-state DC-bus
voltage error is low. Conversely, when Kp and Ki are small,
the effect of real power unbalance on transient performance
is negligible. Therefore, selecting appropriate Kp and Ki
values is crucial in achieving satisfactory control
performance and ensuring that the required power is
supplied to the capacitor.
In the initial approach to setting Kp and Ki values, a trial
and error method is typically used. This involves selecting
initial values for Kp and Ki and then observing the system's
response. Based on the observed behaviour, the values are
adjusted iteratively until the desired performance is
achieved.
3.2 ANN Controller
Artificial Neural networks are mathematical models
designed to process information and makepredictionsbased
on that information. While theseareinspiredbythestructure
and function of biological neural networks,theyarenotexact
replicas.
An Artificial Neural Network (ANN) is a group of simple,
interconnected processingelementsthatcanlearnandadapt.
In a neural network, the individual neurons receive inputs
from other neurons, or from the outside world, and then
process that input before sending an output signal to other
neurons in the network. The weights assigned to each input
determine the strength of its influence on the neuron's
output. The activation function in a neural network is
responsible for calculating the neuron's output based on the
sum of the weighted inputs and a threshold value. This
function plays a critical role in determining the network's
performance and can have a significant impact on the
accuracy of its predictions.
ANNs are identified by their structure, the way they
interact with their surroundings,theirtrainingapproach,and
their information processing capacity. Due to their user-
friendliness, dependability, and fault tolerance, ANNs have
becomea feasible meansofcontrol.Neuralcontrollers,which
share the goal of upgrading hard controllers with intelligent
ones to enhance control quality, are often used as an
alternative to fuzzy controllers.
A feed-forwardneuralnetworkfunctionsasageneratorof
a compensation signal, with the output of the compensator
being determined by the input and its progression over time.
The neural network is trained to produce fundamental
reference currents, which are then compared to obtain
switching signals in a hysteresis band current controller.
3.3 ANFIS Controller
ANFISstandsforAdaptiveNeuro-FuzzyInferenceSystem,
which is a type of artificial neural network thatintegratesthe
strengths of both neural networksandfuzzylogic.ANFISwas
developed by Jang in 1993 as a means of modelling complex
systems that are difficult to describe mathematically.
ANFIS networks use a setofrulesthatcombinefuzzylogic
with neural network training techniques to generate an
output. These rules consist of fuzzy if-then statements that
define the relationship between the inputsandoutputsofthe
system. The ANFIS network uses these rules to generateaset
of membership functions that describe the input-output
relationship.
These membership functions are then used to generate a
set of parameters that are optimized using a neural network
training algorithm. This algorithm adjusts the membership
functions and their parameters to minimize the error
between the network output and the desired output.
ANFIS networks have been applied in a variety of
applications, including control systems, forecasting, data
classification, and image processing. They are especially
useful when the underlying system iscomplexanddifficultto
model mathematically, but there is enough data available to
train the ANFIS network.

4. MODELLING AND SIMULATION
4.1 Modelling of 16 Bus distribution test system
This section illustrates the process of modeling and
implementing the 16-bus test system. The IEEE 16-bus
distribution system was originally proposedbyS.Civnlar[1].
Below is the single-line diagram of the system and the
corresponding bus data.
Fig. 2: 16 Bus Distribution System Single Line Diagram
Line
No
From
Bus i
To
Bus
j
R(ohms)
P.U
X(ohms)
P.U
End Bus Load
P Q
1 1 4 0.075 0.10 2.0 1.6
3 4 5 0.08 0.11 3.0 0.4
2 4 6 0.09 0.18 2.0 -0.4
5 6 7 0.04 0.04 1.5 1.2
7 2 8 0.11 0.11 4.0 2.7
8 8 9 0.08 0.11 5.0 1.8
9 8 10 0.11 0.11 1.0 0.9
6 9 11 0.11 0.11 0.6 -0.5
10 9 12 0.08 0.11 4.5 -1.7
15 3 13 0.11 0.11 1.0 0.9
14 13 14 0.09 0.12 1.0 -1.1
16 13 15 0.08 0.11 1.0 0.9
12 15 16 0.04 0.04 2.1 -0.8
4 5 11 0.04 0.04
13 10 14 0.04 0.04
11 7 16 0.12 0.12
Table 1: Bus Data of 16 Bus Distribution System
4.2 Modelling of UPQC
The study assesses a simplified control algorithm for
UPQC byanalyzingsimulationoutcomesfromSIMULINK.The
table below presents the parameters of the simulated UPQC
system proposed by the study.
System Parameters
Supply Voltage Vabc 23KV
Supply frequency fs 50Hz
DC Link Voltage Vdc 700V
Table 2: UPQC Parameters
Fig. 3: UPQC Simulation Diagram
4.3 Modelling of Series Active Filter
The series active filter was modelled as described in the
series control.
Fig. 4: SEAF Simulation Diagram
4.4 Modelling of Shunt Active Filter
For modeling of shunt activefilter,thedirectvoltageerror
manipulation approach of the DC link is used. For error
calculation between DC link actual value andreferencevalue,
the PI controller is used and the Ploss calculated by the
controller is fed into compensating current calculation block
where compensating current is calculated by means of
formulae written according to PQ theory. The reference
current generated is fed intothehysteresisbandcontrollerto
generate gating pulse for 3phase current source inverterand
the resulting current is fed using a coupling inductor.

Fig. 5: SHAF Simulation Diagram
4.5 Modelling of PI Controller
Traditionally, a PI controller is used to regulate the DC
link voltage in the UnifiedPowerQualityConditioner(UPQC).
However, the effectiveness of this controller depends on
system parameters. The controller's output is then directed
to the shunt inverter's current control system to ensure that
the required amount of active power is drawn from the grid
to maintain the DC capacitor voltage. To achieve optimal
performance of the UPQC under dynamic power system
conditions, such as voltage fluctuations, load changes, and
unbalanced loads, the DC link voltage controlresponseneeds
to be rapid with minimal delay time and lower overshoot. In
the simulation, the Kp and Ki values used for the controller
are 10 and 0.1, respectively. [7]
Fig. 6: PI Controller Simulation Diagram
4.6 Modelling of ANN Controller
Neural network training was initiated using the fitting
app. Input data and output data were mapped and 10
neurons were selected to train using the Levenberg-
Marquardt training algorithm. The function-fitting neural
network is generated and deployed in the model.
Fig. 7: ANN Controller Simulation Diagram
4.7 Modelling of ANFIS Controller
The input-output data from the PI controller was
imported to the neuro-fuzzy designer and the fuzzy file was
generated using the grid partition method. The generated
fuzzy was optimized using the hybrid method and the
number of iterations was selected and the ANFIS model was
trained. With each iteration, the error value was optimized.
The generated .fiz file is implemented in a fuzzy logic
controller and deployed in the model.
Fig. 8: ANFIS Controller Simulation Diagram
5. RESULTS AND ANALYSIS
The 16-bus distribution system is modeled in
MATLAB SIMULINK software and the power quality is
conditioned by using UPQC at the 16th Bus, with various
controllers like PI controller, ANN controller, and ANFIS
controller and the outputs have been compared.

Fig. 9: Simulation for 16 bus distribution system
Fig. 10: Bus 16 Voltage waveform without UPQC
Fig. 11: Bus 16 Voltage waveform with UPQC
Tocalculatetotalharmonicdistortion(THD),a16-bussystem
was modeled with a nonlinear load consisting of a diode
rectifier and RC load. The load's resistance (R) was set to 10
ohms, and its capacitance (C) wassetto500microfarads.The
results of the FFT analysis are shown below.
o
Fig. 12: Simulation for 16 bus distribution system with
Non-Linear Load
Fig. 13: Load current waveform without UPQC
Fig 14: THD of Load current with PI UPQC

Fig. 15: THD of load current using ANN UPQC
Fig. 16: THD of load current using ANFIS UPQC
THD Comparison of Load Current( fmax=2500Hz)
Without
UPQC
PI Controller ANN Controller ANFIS
Controller
9.89% 6.26% 5.85% 5.56%
Table 3: Load Current THD
THD Comparison of Load Voltage( fmax=2500Hz)
Without
UPQC
PI Controller ANN Controller ANFIS
Controller
29.25 1.76% 1.70% 1.69%
Table 4: Load Voltage THD
6. CONCLUSION
The Unified Power Quality Conditioner (UPQC) is a device
that can effectively address a variety of powerqualityissues,
including voltagefluctuationsandharmonics,reactivepower
problems, and poor power factors. This study focuses onthe
necessary system configuration and control topologyforthe
UPQC, as well as a proposed control technique utilizing PI,
ANN, and ANFIS controllers individually. Experimental
results demonstrate that both ANN and ANFIS controller
exhibit similar levels of total harmonic distortion (THD), but
the ANFIS-based UPQC displays better dynamic response
and fast learning and achieves a THD that is 11% lower than
that of the PI controller.
REFERENCES
[1] S. Civanlar, J. J. Grainger, H. Yin and S. S. H. Lee,
"Distribution feeder reconfiguration for lossreduction,"
in IEEE Transactions on Power Delivery, vol. 3, no. 3, pp.
1217-1223, July 1988
[2] Singh, Manish Kumar, and Vivek Saxena. "Voltage
conditioning and harmonic mitigation using UPQC: A
review." Innovations in Cyber Physical Systems: Select
Proceedings of ICICPS, 2021
[3] S.Thirukkovai, J.Venkatesan and Dr.S.M.Girirajkumar
“Voltage Sag/Swell Mitigation Using UPQC”
International Journal of Scientific & Engineering
Research, Volume 5, Issue 4, April-2014
[4] Swaroopa S. Bhosale, Y.N. Bhosale, Uma M. Chavan and
Sachin A. Malvekar “Power Quality Improvement by
Using UPQC: A Review” International Conference on
Control, Power, Communication and Computing
Technologies (ICCPCCT), 2018
[5] Nikita Gupta and K. Seethalekshmi“Acomparativestudy
of different control algorithm used in unified power
quality conditioner for power qualityimprovement”Int.
J. Intelligence and Sustainable Computing, Vol. 1, No. 2,
2021.
[6] Hoon, Y.; Mohd Radzi, M.A.; Hassan, M.K.; Mailah, N.F.
Control Algorithms of Shunt Active Power Filter for
Harmonics Mitigation: A Review. Energies 2017
[7] Mohammed Nizar, SarmilaHar Beagam, Jayashree.
“COMPARISION OF PI CONTROLLER & FUZZY LOGIC
CONTROLLER USING UNIFIED POWER QUALITY
CONDITIONER ” International Research Journal of
Engineering and Technology (IRJET) Volume: 03 Issue:
06, June 2016
[8] L. H. Tey, P. L. So and Y. C. Chu, "Neural network-
controlled unified power quality conditioner for system
harmonics compensation," IEEE/PES Transmission and
DistributionConferenceand Exhibition,Yokohama,Japan,
2002
[9] Manivasagam, Rajendran, and Rajendran Prabakaran.
"Power quality improvement by UPQC using ANFIS-
based hysteresis controller." International Journal of
Operational Research 37.2, 2020

Comparative Analysis of Power Quality Enhancement of Distribution System using Unified Power Quality Conditioner

More Related Content

Similar to Comparative Analysis of Power Quality Enhancement of Distribution System using Unified Power Quality Conditioner (20)

More from IRJET Journal (20)

Recently uploaded (20)

Comparative Analysis of Power Quality Enhancement of Distribution System using Unified Power Quality Conditioner