85 CHAPTER 4 ENHANCING VOLTAGE STABILITY AND LOSS MINIMIZATION USING ANN AND PSO WITH UPFC 4.1 INTRODUCTION One of the important operating requirements of a reliable power system is to maintain the voltage to ensure a high quality of customer service. The main reason for the stability problem occurring in the system is due to the sudden increase in load power as well as any fault that occurs in the system. A genetic algorithm is used to optimize the various process parameters involving FACTS devices in a power system. The various parameters taken into consideration are the location of the device, their type and their rated value of the devices as in Nikoukar and Jazaeri (2007). The optimal location and size of SVC device for decreasing voltage stability index, power loss and voltage deviation use PSO and GA for different loading condition as in Kalivani and Kamaraj (2012). An application of Bacterial Foraging (BF) algorithm in optimizing the optimal location and design of Thyristor controlled series capacitor for voltage profile improvement and minimization of losses in a power system utilized the TCSC as the control variable as in Senthilkumar and Renuga (2010). Optimal location and setting of SVC and TCSC devices use Non- dominated Sorting Practical Swarm Optimization (NSPSO) for voltage stability enhancement as in Benabid et al (2009). Power loss minimization
37
Embed
CHAPTER 4 ENHANCING VOLTAGE STABILITY AND LOSS ...shodhganga.inflibnet.ac.in/bitstream/10603/17396/9/09_chapter 4.pdf · ENHANCING VOLTAGE STABILITY AND LOSS MINIMIZATION USING ANN
This document is posted to help you gain knowledge. Please leave a comment to let me know what you think about it! Share it to your friends and learn new things together.
Transcript
85
CHAPTER 4
ENHANCING VOLTAGE STABILITY AND LOSS
MINIMIZATION USING ANN AND PSO WITH UPFC
4.1 INTRODUCTION
One of the important operating requirements of a reliable power
system is to maintain the voltage to ensure a high quality of customer service.
The main reason for the stability problem occurring in the system is due to the
sudden increase in load power as well as any fault that occurs in the system. A
genetic algorithm is used to optimize the various process parameters
involving FACTS devices in a power system. The various parameters taken
into consideration are the location of the device, their type and their rated
value of the devices as in Nikoukar and Jazaeri (2007).
The optimal location and size of SVC device for decreasing voltage
stability index, power loss and voltage deviation use PSO and GA for
different loading condition as in Kalivani and Kamaraj (2012). An application
of Bacterial Foraging (BF) algorithm in optimizing the optimal location and
design of Thyristor controlled series capacitor for voltage profile
improvement and minimization of losses in a power system utilized the TCSC
as the control variable as in Senthilkumar and Renuga (2010).
Optimal location and setting of SVC and TCSC devices use Non-
dominated Sorting Practical Swarm Optimization (NSPSO) for voltage
stability enhancement as in Benabid et al (2009). Power loss minimization
86
uses Fuzzy multi-objective formulation and Genetic algorithm using TCSC as
in Bagriyanik et al (2003). Utility of PSO for loss minimization and
enhancement of voltage profile use UPFC as in Kannan and Kayalvizhi
(2011).
PSO techniques for optimal location and parameter setting of TCSC
to improve the power transfer capability, reduce active power losses, improve
stabilities of the power network and decrease the cost of power production
and fulfill the other control requirements by controlling the power flow in
multi-machine power system network as in Rashed et al (2007). PSO based
algorithm uses optimal Automatic Voltage Regulator (AVR) values, On-Line
Tap Changer (OLTC) settings and minimum number of Reactive Power
Compensation Equipment (RPCE). These minimize the real power losses,
with a view to improve the voltage stability of the system as in Ajay-D-
Vimalraj et al (2008).
To maintain the stability of the system, finding the location for
fixing the FACTS controller and also the amount of voltage and angle to be
injected is more important. The hybrid technique includes Artificial Neural
Network and Particle Swarm Optimization. The proposed hybrid method
using MATLAB for different systems such as IEEE 14, 30, 57 and 118 bus
systems has been carried out and the performance evaluation is presented.
4.2 POWER FLOW CALCULATION USING NEWTON-
RAPHSON METHOD
There are different methods available for load flow calculation.
Among them the Newton-Raphson method is the most commonly used
(explained in section, 3.3. By using Newton-Raphson load flow analysis
method, the real and reactive power flow in the bus are calculated using the
Equations (3.6) and (3.7) respectively. Using the equation the real and
87
reactive power flow between the buses are computed. The overall process
takes place in the proposed method as shown in Figure 4.1.
Figure 4.1 Overall process takes place in the proposed method
Start
Computing Power flow between the buses using Newton Raphson method
UPFC Modelling
Training ANN using generated dataset
Training ANN to obtain the optimal location for fixing UPFC
PSO to compute the amount of voltage and angle to be injected to reduce power loss
Stop
Generating training dataset for fixing UPFC
88
4.3 UPFC MODELLING
A combination of Static Synchronous Compensator (STATCOM)
and Static Synchronous Series Compensator (SSSC) which are coupled via a
common DC link, allows bidirectional flow of real power between the series
output terminals of the SSSC and the shunt output terminals of the
STATCOM. They are controlled to provide concurrent real and reactive series
line compensation without an external electric energy source. The UPFC, by
means of angularly unconstrained series voltage injection, is able to control,
concurrently or selectively, the transmission line voltage, impedance and
angle or alternatively, the real and reactive power flow in the line. The UPFC
also provides independently controllable shunt reactive compensation. A
detailed explanation is given in section 3.4.
The power injection in terms of voltage and power angle, due to the
insertion of UPFC is given in the Equations (3.8) to (3.11). The amount of
voltage and angle to be injected and the optimal location for fixing UPFC in
the system are identified using the hybrid technique. The next step is
identifying the location for fixing UPFC using artificial neural network.
4.4 ANN TO OBTAIN LOCATION FOR FIXING UPFC
In the proposed method, ANN is used to identify the optimal
location for fixing UPFC to maintain the stability of the system. ANN
consists of two stages: training stage and testing stage. In the training stage,
neural network is trained based on the training dataset and in the testing stage,
if the input variable is given; it gives the corresponding output variables. A
detailed explanation is given in section 3.5.1 and 3.5.2.
89
4.5 PARTICLE SWARM OPTIMIZATION
Particle swarm optimization is a relatively new evolutionary
algorithm that may be used to find optimal (or near optimal) solutions to
numerical and qualitative problems. Particle swarm optimization was
originally developed by Kennedy and Eberhart (1995) and emerged from
earlier experiments with algorithms that modelled the flocking behaviour seen
in many species of birds. In simulations, birds would begin by flying around
with no particular destination and spontaneously formed flocks until one of
the birds flew over the roosting area. Due to the simple rules, the birds used to
set their directions and velocities. A bird pulling away from the flock, in order
to land at the roost, would result in nearby birds moving towards the roost.
Once these birds discovered the roost, they would land there, pulling more
birds towards it and so on until the entire flock had landed. Finding a roost is
analogous to finding a solution in a field of possible solutions in a solution
space.
The manner in which a bird which has found the roost leads its
neighbours to move towards it, increases the chances that they will also find
it. This is known as the socio-cognitive view of mind. The socio-cognitive
view of mind means that a particle learns primarily from the success of its
neighbours. The concept of the PSO consists of, at each time step, changing
the velocity of (accelerating) each particle toward its pbest and lbest locations
(local version of PSO). Acceleration is weighted by a random term, with
separate random numbers being generated for acceleration toward pbest and
lbest locations. In the past several years, PSO has been successfully applied in
many research and application areas.
90
4.5.1 Basic Terms used in PSO
The basic terms used in PSO technique are stated and defined as
follows (Abido 2003):
1. Particle X (I): It is a candidate solution represented by a k-dimensional
real-valued vector, where k is the number of optimized parameters. At
iteration i, the jth particle
X (i,j) can be described as:
X i (i ) = [ X j 1 (i ); X j 2 (i );.....X jk (i );.....X jd
where, X’s are the optimized parameters and d represents the number of
control variables
2. Population: It is basically a set of n particles at iteration i.
pop (i )= [ X 1 (i ), X 2 (i ), .........X n (i)]T
where: n represents the number of candidate solutions.
3. Swarm: Swarm may be defined as an apparently disorganized population
of moving particles that tend to cluster together while each particle seems to
be moving in a random direction.
4. Particle velocity V (i): Particle velocity is the velocity of the moving
particles represented by a d-dimensional real valued vector. At iteration i, the
jth particle
Vj (i) can be described as:
V j (i ) = [V j1 (i );V j2 (i );.....V jk (i );.....V jd (i);]
91
where, V jk (i) is the velocity component of the jth particle with respect to the
kth dimension.
5. Inertia weight w (i): It is a control parameter, which is used to control the
impact of the previous velocity on the current velocity. Hence, it influences
the trade-off between the global and local exploration abilities of the particles.
For the initial stages of the search process, large inertia weight to enhance
the global exploration is recommended while it should be reduced at the last
stages for better local exploration. Therefore, the inertia factor decreases
linearly from about 0.9 to 0.4 during a run.
6. Individual best X* (i): As the particles moves through the search space, it
compares its fitness value at the current position to the best fitness value it has
ever reached at any iteration up to the current iteration. The best position that
is associated with the best fitness encountered so far is called the individual
best X* (i).
7. Global best X** (t): Global best is the best position among all the
individual best positions achieved so far.
8. Stopping criteria: Termination of the search process will take place
whenever one of the following criteria is satisfied:
The number of the iterations since the last chance of the best
solution is greater than a pre-specified number.
The number of iterations reaches the maximum allowable
number.
The flow chart of the PSO procedure is shown in Figure 4.2.
92
Figure 4.2 Flow chart of the PSO procedure
Start
Generate an initial particle of voltage and angle nv,2v,1viV & n,2,1i
Evaluation function for each particle
from the equation (4.1)
Check and Update
Personal best and Global best
Success
Stop
CheckStopping criteria
UpdateParticle from the equation (4.2) & (4.3)
I=i+1
No
Yes
93
4.5.2 Computing and using PSO
PSO is used to compute the amount of voltage and angle to be
injected in UPFC. PSO consists of four stages namely, generating initial
particle, evaluation function, updating initial particle and termination.
Generation of initial particle
The initial process in PSO is generation of initial particle. The
initial particle is ni vvvV ,, 21& ni ,, 21
, where nvvv ,...., 21
and n,...., 21 are the particles and n is the number of particles. Consider the
UPFC fixing location to be and . For that location, the voltage and angle
injecting values are computed using PSO. After generating the initial particle,
the next step is to evaluate the initial particle using the evaluation function.
Evaluation function
Evaluation function is used in PSO to identify the best particle from
the set of initialized particle. The evaluation function used in the proposed
method is the total power loss.
N
jiijmm BYjViVConjalfunctionEvaluation
1,**)(*)(Re (4.1)
where, mV is the voltage magnitude, ijY is the Y-bus matrix, and B is the base
MVA value.
94
Updating initial particle
In this stage, the initial particles generated are updated and then the
evaluation function is applied. The particles are updated using the Equation
Table 4.6 depicts the real power loss in the system in normal case,
sudden increase in load power case and after connecting UPFC using the
proposed method.
Figure 4.6 Comparison of real power loss of IEEE 30 bus system for different scenarios
104
From Figure 4.6, it is clear that the real power loss that occurred for
normal load condition is 17.914 MW. After a sudden increase in the load
power in bus 6, the real power loss is increased to 21.565 MW and then by
using the proposed method with UPFC, the real power loss gets reduced to
16.061 MW.
Table 4.7 Comparison of real power loss of IEEE 30 bus system for the other methods
Method Real power loss (MW) Nikoukar and Jazaeri (2007) GA 17.695Kalaivani and Kamaraj (2012) PSO 17.543Ajay-D-Vimalraj et al (2008) PSO 16.8264Proposed method ANN with PSO 16.061
Table 4.7 makes it evident that the proposed method obtains 9.23 %
of loss reduction compared to GA value reported in Nikoukar and Jazaeri
(2007), 9.15 % of loss reduction compared to PSO value mentioned in
Kalaivani and Kamaraj (2012) and 4.55 % of loss reduction compared to PSO
value indicated in Ajay-D-Vimalraj et al (2008). This shows the effectiveness
of the proposed approach which minimizes the real power loss compared to
the other methods.
4.6.3 IEEE 57 Bus System
The test system consists of 7 generator buses (bus no. 1, 2, 3, 6, 8,