International Journal of Advances in Engineering and Management (IJAEM)

Volume 4, Issue 5 May 2022, pp: 763-772 www.ijaem.net ISSN: 2395-5252

Load flow analysis using Forward and

Backward sweep, and minimising power
losses using Genetic Algorithm.
Tanish Deosaria, Srijan Choudhary, Tajender Meena
--------------------------------------------------------------------------------------------------------------------------------------
Submitted: 05-05-2022 Revised: 10-05-2022 Accepted: 13-05-2022
-----------------------------------------------------------------------------------------------------------------------------------
ABSTRACT
This thesis presents a Backward/Forward Sweep I. INTRODUCTION
Method to analyse the load flow in radial In general, the definition of an electric
distribution system. The R/X ratios of distribution power system includes generating, transmission
system having radial structure are high. So the and distribution system. Among these the economic
Newton-Raphson and fast decoupled methods are importance of the Distribution System is very high.
failed with distribution system. On the other hand, The amount of investment involved in distribution
the proposed method is one of the most effective system dictates careful planning, design,
methods for the load-flow analysis of the radial construction and operation. It holds a very specific
distribution system and it is easy to implement and position in the power system since it is the main
there is no need of any complex renumbering of link between bulk power and consumers. It is
branches and nodes, or any matrix calculation. This important to plan distribution network effectively
method uses only the linear equations based on to meet the present growing domestic, industrial
Kirchhoff‟s formulation and is used to determine and commercial load day-by-day. Distribution
power losses for each bus branch and voltage networks have attained an overwhelming research
magnitudes for each node. interest in the academics as well as in the industries
After analysis load flow analysis by forward and community nearly from last 30 years.
backward sweep method we observe there is some
power losses that occur in our distributed system. 1.1DISTRIBUTION SYTEM
So to optimized our system we introduce genetic An electric power distribution system
algorithm for optimal capacitor placement. transmits electricity from transmission system to
The placement of shunt capacitor banks at individual consumers. It is the final phase in the
optimal locations in the distribution network and delivery of electric power. It provides power to
their sizing can effectively reduce the losses in individual customer premises. The objective of
the utility network. It also helps in the maximum distribution system planning is to assure that the
active power flow through the existing growing demand for electricity in terms of
distribution lines which. This also increases the accelerating growth rates and high load densities
power transfer capacity of feeders and improves are often satisfied in an optimum way by additional
the voltage profile of the feeders which leads to distribution Systems from the secondary
reduced investment of transmission network. conductors through the majority power substations,
This paper presents a method for optimal which are both technically adequate and fairly
placement and sizing of the capacitors in radial economical.
distribution feeder using Genetic algorithm (GA) Distribution transformers again decrease
with an objective of loss reduction and voltage the voltage to lower level to the utilization voltage
profile improvement. The results have been of household appliances and typically feed several
validated using MATLAB programming. An customers through secondary distribution lines at
IEEE 33-bus distribution test feeder is employed this voltage. The distribution of electric power to
for testing the proposed algorithm. The optimal consumers is done with a much lower voltage level
sizing of the capacitors has been suggested in as compared to that of the transmission of power
terms of thekVAr. over long distances. It generally consists of feeders,
Keywords:Backward/Forward sweepmethod, circuit breakers and the service mains. Fig. 1.1
Distribution system, Load-flow Analysis.

DOI: 10.35629/5252-0405763772 Impact Factor value 7.429 | ISO 9001: 2008 Certified Journal Page 763
 International Journal of Advances in Engineering and Management (IJAEM)
Volume 4, Issue 5 May 2022, pp: 763-772 www.ijaem.net ISSN: 2395-5252

shows the single line diagram of a typical low

tension distribution system.

Fig. 1.2 Radial Distribution Network

Criteria assumed for node and line numbering in
the network is as follows:
1. The nodes are numbered sequentially in
Fig. 1.1 Elements of Distribution System ascending order proceeding from layer to
layer, in such a way that any path from the root
1.2TYPES OF DISTRIBUTION SYSTEM node to a terminal node encounters nodes
Distribution networks are divided into two types: numbered in the ascending order.
a) Radial Distribution System 2. Each branch starts from the sending bus (root
side) and is identified by the number of its
b) Ring Main Distribution System
(unique) ending bus.
The ring main system is generally more
The main advantage of radial network is
expensive than the radial system because more
that its construction is simple, low initial cost, can
switches and conductors are required in ring main
be used when generation is at low voltage. It is
system leading to high construction cost and it is
preferred when the substation is located at the
not preferred to choose ring main system when the
centre of the load.
generation is at low voltage. Due to these reasons,
1.4FEATURES OF RADIAL DISTRIBUTION
radial system is used in distribution system.
NETWORK
The conductors which are used in distribution
1.3RADIAL DISTRIBUTION SYSTEM
systems are characterized by:
Separate feeders are radiated from one
substation and feed the distributors at only one end. • High R/X ratios
This is termed as radial system. Radial distribution • Uncertainties and imperfections
is the power distribution system in which the power • Radial structure
is delivered from the main branch to the sub • Large number of nodes and branches
branches then it split out from the sub branches • Distributed generation and topology
again as shown in Fig. 1.2, where power is configurations
transferred from the node and it splits at L1.There
are no loops in the radial distribution network and
• Dynamic change in imposed load
1.5LOAD FLOW STUDIES
each bus is connected to the source via exactly one
Load-flow studies are used to make sure
path. This network configuration is the cheapest
that the electrical power transmission, through the
and least reliable. The Radial Distribution System
grid system, from generators to consumers is
is generally used in sparsely populated areas.
stable, reliable and economic. The increasing
presence of distributed alternative energy sources,
frequently in geographically remote locations,
complicates load flow studies and has triggered a
resurgence of interest in the topic. In a three phase
AC power system, active and reactive power flows
from the generating station to the load through
different networks buses and branches. The flow of
active and reactive power is called flow or load
flow. Power flow studies provide a systematic
mathematical approach for computation of various
bus voltages, their phase angle, active and reactive
DOI: 10.35629/5252-0405763772 Impact Factor value 7.429 | ISO 9001: 2008 Certified Journal Page 764
 International Journal of Advances in Engineering and Management (IJAEM)
Volume 4, Issue 5 May 2022, pp: 763-772 www.ijaem.net ISSN: 2395-5252

power flow through different branches, generators each node are collected to establish it. The network
and loads under steady state condition. information can be assembled in one table, as
shown in Table 2.1.
II. THEORY:
Load-flow analysis is a extensively used Node kVA kVAr
tool in transmission system in several applications
such as power generating scheduling and very
useful for many applications in distribution system ……..
……..
such as network analysis, load control, network ……..
reorganization, integration of generation and
integration of electric vehicle. To understand the ……..
…….. ……..
nature of the installed network, Load-flow studies
are performed on power systems. Load-flow is used
…….. …….. ……..
to discover the static performance of the system
and usually uses simplified notation such as a one-
line diagram and per-unit system. It focuses on Table 2.1: Load Data Table
various forms of AC power (voltages, voltage
angles, real power and reactive power). It examines Line data represents network conductor
the power systems in normalsteady-state operation. characteristics; in utilities, line data are grouped in
Due to radial or weakly meshed networks, the GIS, or by knowing the electrical characteristics
high R/X ratios, multi-phase, unbalanced operation, of conductors used, it can be brought together
unbalanced distributed load and distributed manually, the laying mode and the year of the
generation, transmission grid load flow methods or conductor laying. The network information can be
the Jacobian-based methods such as Newton- assembled in one table, as shown in Table 2.2.
Raphson, Gauss-Seidel and fast decoupled methods
are failed with distribution systems. Many Sending nodeReceiving R X (Ohm)
approaches have been developed in the past for node (Ohm)
distribution system load-flow analysis. Due to
computational efficiencies and solution accuracies, ……..
the ladder network theory and the …….. ……..
……..
backward/forward sweep methods are the
commonly used approaches for distribution system …….. ……..
load flow analysis. ……..
……..
In this thesis, standard backward/forward
sweep method is used for radial distribution load
…….. …….. …….. ……..
flow analysis. An initial calculation is made from
the large data to organize the radial distribution
information into a main line and its derivations.
Table 2.2: Line Data Table
After the organization into a suitable model,
voltages of all other nodes are set to nominal
However, the network topology can be changed for
voltage and then an iterative process is carried out
maintenance activities, emergency operations, or
by calculating currents in derivative lines and then
network configurations.
the branch current in the main line is computed in a
backward sweep using KCL. The node voltages are
computed in a forward sweep using the voltage 2.2 FORMULATION OF THE PROBLEM
In a distribution system, the power flows are
drop calculation. This backward and forward sweep
technique is repeated until voltage magnitudes in enumerated by the following set of simplified
recursive equations derived from the single-line
each node in the current iteration and the preceding
diagram, as shown in Fig
iterations are lower than the tolerance limit.

2.1LOAD FLOW ANALYSIS

Analysis of Load-flow requires two
inputs, Line data and Load data, to provide
electrical network characteristics.
In Load data, active power injections and
the consumptions of the active/reactive power at

DOI: 10.35629/5252-0405763772 Impact Factor value 7.429 | ISO 9001: 2008 Certified Journal Page 765
 International Journal of Advances in Engineering and Management (IJAEM)
Volume 4, Issue 5 May 2022, pp: 763-772 www.ijaem.net ISSN: 2395-5252

SENDING END PT,loss(k,k+1) – Total Real power loss in the line

RECEIVING END section
QT,loss(k,k+1) – Total Reactive power loss in the
line section

2.3 PROPOSED LOAD FLOW ANALYSIS

METHOD
Let us consider a radial network, the
Backward/Forward Sweep technique for the
computation of load-flow is an iterative method
within which, at every iteration two processes are
performed, Backward sweep and Forward sweep.
The load flow of one source network will be
resolved iteratively from two sets of algorithmic
equations. The first set of equations for calculations
of the load flow through the branches initiating
from the last branch and continuing in the
backward direction towards the root node. The
Single Line Diagram other set of equations are for calculating the voltage
magnitude and the angle of each node initiating
The power flow can be used to obtain the voltage from the root node and continuing in the forward
magnitude, power losses of the radial distribution direction towards the last node.
network. The objective function is to find the In Fig. 2.2, the diagram of the general principle of
power flow. the proposed load flow method is given.

where,
Pk – Real power flowing out of bus
Qk– Reactive power flowing out of bus
PLk+1 – Real load power at bus k+1
QLk+1 - Reactive load power at bus k+1
The power loss in the line section connecting buses
k and k+1 can be computed by the formulas given
below.

where, Fig. 2.2 Diagram of the Proposed Load Flow

Ploss(k,k+1) – Real power loss in the line section Analysis
connecting buses k and k+1
Qloss(k,k+1) – Reactive power loss in the line BACKWARD SWEEP: It is a current or load flow
section connecting buses k and k+1 solution with possible voltage updates. Branch
The total power loss of the feeder, PT,LOSS may then current and load flows are updated beginning from
be determined by summing up the losses of all line the branches in the last layer and moving towards
sections of the feeder, which is given as the branches connected to the root node. The
updated effective load flows in each branch are
calculated in the backward propagation
computation considering the node voltages of
previous iteration. This indicates that the voltage
values calculated in the forward method are held
where, constant during the backward propagation and

DOI: 10.35629/5252-0405763772 Impact Factor value 7.429 | ISO 9001: 2008 Certified Journal Page 766
 International Journal of Advances in Engineering and Management (IJAEM)
Volume 4, Issue 5 May 2022, pp: 763-772 www.ijaem.net ISSN: 2395-5252

updated power loads in each branch are transmitted Step 2: Iteration Count Initialization, K = 1
backward along the feeder using backward method. Step 3: Load Current computation
This means that the backward propagation begin at
the extreme end node and proceeds towards the Ij(k) = conj( for j = 2, 3, …..N
source node. (2.8)
FORWARD SWEEP: It is basically a voltage-drop Step 4: Backward Sweep
calculation with possible current and power flow Imn
updates. In a forward sweep, Nodal voltages are
updated initiating from branches in the first layer
towards those in the last. The aim of the forwards (2.9)
propagation is to calculate the voltages at each Step 5: Forward Sweep
node beginning from the feeder source node. The Vn(k) = Vm(k) + Zmn.Imn(k) for all n = 2,
voltage of feeder substation is set at its actual 3,….,N (2.10)
value. During the forward propagation the effective
power in each branch is held fixed to the value Step 6: Error
calculated in backward propagation. Ej(k) = |Vj(k)- Vj(k-1)| for j = 2,
The Backward/Forward sweep algorithm 3,…..,N (2.11)
includes two steps: the backward sweep and the Step 7: Maximum Error
forward sweep. Voltage and currents are computed Emax(k) = max (E2(k), E3(k),
in backward sweep using Kirchhoff‟s Voltage Law E4(k),….., En(k))
and Kirchhoff‟s Current Law from the End node to Step 8: If Emax(k) is less than or equal to tolerance
the Source node. In Forward sweep, the (ε), then the load flow is converging.
downstream voltage is computed starting from the Else update the iteration count to k=k+1 and go to
source node. The input data of this algorithm is step 3 and repeat the steps.
given by node-branch aligned data. Basic data In the proposed load flow method, power
required are the active and reactive powers and summation is done in the backward direction and
classification for sending and receiving nodes. voltages are calculated in the forward direction.
Given below are the major steps of the proposed Figgives the detailed operation of the load flow
algorithm with appropriate equations. calculation using backward forward sweep
algorithm.

Fig. 2.3 Bus Network for the load flow calculation

using Backward/Forward sweep method

Let us assume the radial distribution bus network

given in Fig. 2.3 where N is the total number
nodes.
Step 1: Initialization of Voltages
Vj(0) = Vs 0 for j = 2, 3,…..,N FigOperation of the Load Flow calculation using
(2.7) Backward/forward sweep method
DOI: 10.35629/5252-0405763772 Impact Factor value 7.429 | ISO 9001: 2008 Certified Journal Page 767
 International Journal of Advances in Engineering and Management (IJAEM)
Volume 4, Issue 5 May 2022, pp: 763-772 www.ijaem.net ISSN: 2395-5252

PROBLEM FORMULATION: well as placement of shunt capacitor banks is a

Main objective of capacitor placement is key aspect for the powersystem.
to minimize the losses and improve voltage Objective function
profile in the power system network. To achieve The proposed objective function considers the
this objective, an objective function is balanced three-phase system with time invariant
formulated. Main goal of this objective function loads. Mathematically, the proposed objective
is to minimize the losses of the active and function of the problem is based on the
reactive powers along with the improvement of minimization of the loss in power network and
voltage profile. The section details related to the deviation in the voltages from their pre-defined
load and capacitor model, constraints standard values. The definition of the n objective

formulation, objective function and power loss function is given below. 3 i

calculations are described in the F W1 Ploss  W 2  Q lossW (1 v ) (3) 2
followingsubsections.
where W1, W2 and W3 are weights utilized in
A. Load and capacitormodel
the objective function corresponding to active
The loads and capacitors are modeled as
power loss, reactive power loss and voltage
impedance. The impedance model of loads and
deviation. Ploss is total active power loss andQloss is
capacitors are given by Eq. (1) and Eq. (2)
total reactive power loss in distribution network. vi
respectively.
is voltage magnitude at ith bus of the test system.
ZLoadRLoad jXload (1)
i C. Constraints
i = 1,2,3, ,NL The constraints are another significant
Where NL = loadnumbers part of the optimization model along with the
Z = impedance of ithload objective function. These are required to be
Load
R = resistance of ithload defined in terms of the maximum and minimum
Load
X Load= reactance of ithload limits. The constraints considered in this
i
k = 1,2,3, ,NC research are of inequality type. The main
Where NC = capacitorsnumbers constraints for the voltage are considered in the
Z = impedance of kth capacitor proposed research. The limits of voltage
C
k magnitude at each bus are defined by
X = reactance of kthcapacitor the following expression.
In C the power network, the capacitive
impedance cancels out the effect of inductive D. Calculation of Power Loss
reactance and helps in minimization of the losses The complex power at the ith bus is given by the
and improvement in the voltage profile. relation
Excessive capacitive reactive power will results P jQV *I
to the leading power factor which is an
undesirable phenomenon in the power system as Where,
this condition produces the heating losses in the Pi : Load active power
end user equipments. Hence, optimal sizing as Qi : Load reactive power
Vi : Voltage at ith bus

i
I : Load current at ith bus
The Gauss-Seidel iterative method is used to calculate the bus
voltage and line losses

DOI: 10.35629/5252-0405763772 Impact Factor value 7.429 | ISO 9001: 2008 Certified Journal Page 768
 International Journal of Advances in Engineering and Management (IJAEM)
Volume 4, Issue 5 May 2022, pp: 763-772 www.ijaem.net ISSN: 2395-5252

GENETIC ALGORITHM: Generation: At each iteration, the

This section details the theoretical genetic operators are used to perform a series of
concepts of the genetic algorithm. The computations on the current population to
foundations for genetic algorithms were first produce a new population. The successive
introduced by John Holland [28] and then populations produced are known as
elaborated in detail using the tutorials by David newgeneration.
Goldberg [29]. The genetic algorithm is a global Parents and children: To create the next
search technique that can be utilized to solvethe generation, the GA selects a number of
optimizationproblems.Thisisbasedonthetheoryof individuals in the existing population which is
natural selection and biological evolution called parents. This generation is used to create
process. GAconsists of population of binary individuals in the next generation known
string, which searches many peaks in parallel aschildren.
[30]. The important terminologies utilized The following genetic operators are
related to the GA are detailed below[31]. applied on parents to form children for next
Individual: The point for whichan generation:
objective function is implemented can be treated Reproduction: This selects the fittest
as an individual. This is a set of values of individuals in the current population to be used
variables for which function is to be optimized. in generating the next population. These children
The value of objective function for an individual are known as Elitechildren.
is called its score. The vector entries related to Crossover: This causes pairs of
the genes of a genome are considered individuals to exchange genetic information
asindividuals. between each other. These children are known as
Population: Population is an array of crossoverchildren.
individuals. If we consider the size of population Mutation: This causes individual genetic
as 100 and number of variables in the objective representations required to be changed following
function as 3, then population can be represented a set of some probabilistic rules. These children
by a matrix of the size 100x3 in which each row are called mutation children.
corresponds to anindividual.

DOI: 10.35629/5252-0405763772 Impact Factor value 7.429 | ISO 9001: 2008 Certified Journal Page 769
 International Journal of Advances in Engineering and Management (IJAEM)
Volume 4, Issue 5 May 2022, pp: 763-772 www.ijaem.net ISSN: 2395-5252

IMPLEMENTATION OF GA FOR CAPACITOR

PLACEMENT
The objective function formulated for III. RESULTS AND DISCUSSION
the optimal placement and sizing of the The proposed Backward/Forward sweep
capacitors is minimized with the help of an algorithm is tested on IEEE 33–bus radial
optimization tool based on GA. The application distribution system using MATLAB coding. IEEE
of GA is iterated number of times taking 33 bus radial distribution network consists of 33
different values of W 1, W2 and W3 to obtain nodes and 32 branches as shown in Fig. 3.1. The
optimum results. The detailed procedure of the base voltage for this system is 11 kV and base
genetic algorithm utilized for implementation of MVA is 100.
capacitors in test network for minimization of
loss and voltage profile improvement.
Generate randomly constructed
solutions (strings) as initial population of
capacitors of given values which are to be placed
at random nodes in the power distribution feeder
i.e. IEEE 13 bus testsystem.
The crossover and mutation operators
are used to obtain new solutions (capacitor
values) using the initial population during
geneticcycle.
The objective function values (power
loss) are estimated for each new solution and IEEE 33-bus Radial Distribution Network
decoded accordingly. The better solution can be
obtained with the help of selectionprocedure. The tolerance is 10-5p.u. and the total
The worst solution is discarded and the number of iterations required is 2. The Voltage
better solution joins the newpopulation. Magnitude (p.u.), Phase Angle (degrees), Real
Individuals in the initial population with (kW) and Reactive (kVAR) Line losses in each
higher ranking in view of fitness value are used branch is given in Table 3.1. Voltage profile of this
to replenish the shrunkenpopulation. system is shown in Fig. The Total Real power loss
A new genetic cycle is started in each iteration is 2.066303e02 KW.
till the termination criterion is achieved which is The Total Reactive power loss is 1.378083e02 KW.
themaximum number ofgenerations. The proposed genetic algorithm-based
The step-by-step procedure for capacitor allocation solution methodology for the optimal placement
using the proposed algorithm is described in the of capacitors in the power distribution network
flow chart of the Fig. 1. The size of capacitor at a has been implemented on 33-bus IEEE test
particular bus location of the test system is used as

decision variable. The encoding strategy of each feeder. The candidate places for capacitor
individual is used which forms a possible solution. placement and quantity of reactive compensation

DOI: 10.35629/5252-0405763772 Impact Factor value 7.429 | ISO 9001: 2008 Certified Journal Page 770
 International Journal of Advances in Engineering and Management (IJAEM)
Volume 4, Issue 5 May 2022, pp: 763-772 www.ijaem.net ISSN: 2395-5252

are decided by GA and tested for losses of the 14 0.8874 0.9356

active and reactive powers by load flow analysis 15 0.8856 0.9342
based on Gauss-Seidel method. The losses of the 16 0.8828 0.9329
active and reactive powers in the test network 17 0.8820 0.9309
without reactive power compensation and 18 0.9953 0.9303
optimal reactive power compensation using 19 0.9906 0.9969
genetic algorithm are provided in Table 4. The 20 0.9896 0.9934
active power losses decrease by 66.51% and
21 0.9888 0.9927
reactive power losses decreases by 60.81% of
22 0.9722 0.9920
original values.
23 0.9632 0.9825
TABLE IV. RESULTS RELATED TO LOSSES 24 0.9588 0.9758
OF ACTIVE AND REACTIVE POWERS 25 0.9292 0.9725
Loss of Loss of 26 0.9257 0.9629
active reactive 27 0.9101 0.9615
power power 28 0.8989 0.9568
Before Optimization 206.73 29 0.8941 0.9537
137.90 30 0.8884 0.9521
After optimization by 141.842 96.4364 31 0.8871 0.9481
GA 32 0.8867 0.9472

The size of shunt capacitors is taken as

discrete values of the capacity 50 kVAr which is
generally used in the 11 kV distribution systems.
The calculated values of reactive power
compensation are converted in to nearest
capacity as integral multiple of minimum
available capacity. The bus voltages before and
after compensation as well as capacity of
capacitor units at candidate buses are provided in
the Table 5. All voltages after optimal capacitor
placement are found to be within permissible
limits as specified.

RESULTS OF BUS VOLTAGES AND

REACTIVE COMPENSATION
Bus Bus Voltages Bus voltages after
No. without Reactive optimal reactive
compensation compensation IV. CONCLUSIONS:
using A new method for solving the load flow
GA problem for radial distribution feeders without
0 1.0000 1.0000 using conventional load flow methods like Gauss
Seidel, Newton-Raphson, Fast Decoupled methods
1 0.9960 0.9975
is presented in this thesis. This algorithm uses
2 0.9770 0.9860
simple algebraic equations to calculate repeatedly
3 0.9668 0.9805 the outgoing powers and voltage magnitudes of
4 0.9568 0.9805 different nodes and discrepancy at the last nodes of
5 0.9318 0.9752 main feeder and laterals and depending upon this
6 0.9271 0.9641 discrepancy the substation injection is corrected
7 0.9205 0.9625 cautiously, and this process is repeated until
8 0.9119 0.9529 convergence. This makes the algorithm very strong
9 0.9040 0.9490 and numerically efficient for convergence for wide
10 0.9028 0.9456 alternation of distribution network. The results for
11 0.9008 0.9449 IEEE 33-bus test system have been tabulated in
12 0.8924 0.9438 Table 3.1and voltage profile is shown in Fig.3.2
13 0.8894 0.9378 and it was found that the proposed load flow

DOI: 10.35629/5252-0405763772 Impact Factor value 7.429 | ISO 9001: 2008 Certified Journal Page 771
 International Journal of Advances in Engineering and Management (IJAEM)
Volume 4, Issue 5 May 2022, pp: 763-772 www.ijaem.net ISSN: 2395-5252

method is suitable for fast convergence Ching-Lien Huang, “Solving the capacitor
characteristics, high R/X ratios and radial structure. placement in a radial distribution system
This paper proposes a solution for using Tabu Search approach,” IEEE
placement of shunt capacitor banks in the power Transactions on Power System, vol. 11, no.
distribution network which can be easily 4, Nov.1996
implemented in the practical power system [9]. Bompard, iE. iCarpaneto,” iConvergence iof
networks. The proposed method is based on the ithe ibackward/forward isweep imethod ifor
genetic algorithm. The algorithm is implemented ithe iload-flowanalysis iof iradial
on 33-bus IEEE distribution test system. The idistribution isystems” iElectrical iPower
optimal location and sizes of shunt capacitors is iand iEnergy iSystems i22 i(2000) i521–530.
obtained for minimum loss in the network and i
optimal voltage profile. The results have been [10]. Goswami, iS.K iand iBASU, iS.K.: i„Direct
validated in the MATLAB. The developed isolution iof idistribution isystems‟,IEE
algorithm can also be applied on real network of iProc. iC. i, i188, i(I),pp. i78-88, i1991. i
large electricity utilitynetwork. [11]. G.B. iJasmon, iL.H.C. iLee, iDistribution
iNetwork iReduction ifor iVoltage iStability
REFERENCES: iAnalysis iand iLoad iFlow iCalculations,
[1]. S.C.Tripathy, G.Durga Prasad, O.P.Malik iElectrical iPower i& iEnergy iSystems,
and G.S.Hope, “Load Flow for iVol.13,no:1,pp. i9-13, i1991. i
IllConditionedPower Systems by a Newton [12]. Chiang, iH.D.: i„A idecoupled iload iflow
like Method”, IEEE Trans., PAS-101, imethod ifor idistribution ipower inetwork
October 1982, pp.3648-365. ialgorithms, ianalysis iand iconvergence
[2]. M.E. Baran, F.F. Wu, Optimal Sizing of istudy”, iElectrical iPower iand iEnergy
Capacitors Placed on a Radial iSystems, i13 i(3), i130-138, i1991. i
DistributionSystem, IEEE Transactions on [13]. S.Ganesh, i“Network ireconfiguration iof
Power Delivery, Vol.4, no:1, pp.735-743, idistribution isystem iusing iartificial ibee
1989. icolony ialgorithm”, iWASET, iInternational
[3]. B. Stott, “Review of Load-Flow Calculation iJournal iof iElectrical, iElectronic iScience
Methods”, Proceedings of the IEEE, Vol. 62, iand iEngineering, iVol:8No:2, i2014. i
No. 7, July 1974, pp. 916-929.
[4]. R. SrinivasaRao, K. Ravindra,“Power Loss
Minimization in Distribution System Using
Network Reconfiguration in the Presence of
Distributed Generation” IEEE Transactions
On Power Systems, Vol. 28, No. 1, February
2013.
[5]. R. Srinivasas Rao, S. V. L. Narasimham, and
Ramalingaraju, “Optimal capacitor
placement in a radial distribution system
using plant growth simulation algorithm,”
International Journal of Electrical Power and
Energy Systems, vol. 33, pp. 1133-
1139,2011.
[6]. G. Levitin, A. Kalyuzhny, A. Shenkman,
and M. Chertkov, “Optimal capacitor
allocation in distribution systems using a
genetic algorithm and a fast energy loss
computation technique,” IEEE Transactions
on Power Delivery, vol. 15, no. 2, pp. 623-
628, April2000.
[7]. Schmill J. V., “Optimum size and location of
shunt capacitors on distribution feeders,”
IEEE Transactions on Power Systems
Apparatus and Systems, vol. 84, no. 9,
pp.825-832, Sept.1965
[8]. Yann-Chang Huang, Hong-Tzer Yang, and

DOI: 10.35629/5252-0405763772 Impact Factor value 7.429 | ISO 9001: 2008 Certified Journal Page 772