1-s2.0-S0142061515005827-main.pdf

8/19/2019 1-s2.0-S0142061515005827-main.pdf

1/15

A novel method based on adaptive cuckoo search for optimal network

reconfiguration and distributed generation allocation in distribution

network

Thuan Thanh Nguyen a,b,⇑, Anh Viet Truong a, Tuan Anh Phung c

a Faculty of Electrical and Electronics Engineering, HCMC University of Technology and Education, 1 Vo Van Ngan Str., Thu Duc Dist., Ho Chi Minh City, Viet Namb Dong An Polytechnic, 30/4 Str., Di An Dist., Binh Duong Province, Viet Namc Ha Noi University of Technology, 1 Dai Co Viet Str., Ha Noi, Viet Nam

a r t i c l e i n f o

Article history:

Received 4 September 2015

Received in revised form 3 December 2015

Accepted 17 December 2015

Keywords:

Distribution network reconfiguration

Distributed generator

Adaptive cuckoo search algorithm

Power loss reduction

Voltage stability enhancement

a b s t r a c t

This paper proposes a new methodology to optimize network topology and placement of distributed gen-

eration (DG) in distribution network with an objective of reduction real power loss and voltage stability

enhancement. A meta-heuristic cuckoo search algorithm (CSA) inspired from the obligate brood para-

sitism of some cuckoo species which lay their eggs in the nests of other birds of other species for solving

optimization problems is adapted to simultaneously reconfigure and identify the optimal location and

size of DG units in a distribution network. The graph theory is used to determine the search space which

reduces infeasible network configurations of reconfiguration process and check the radial constraint of

each configuration of distribution network. The effectiveness of the proposed method has been validated

on three different distribution network systems at seven different scenarios. The obtained results show

well the effectiveness and performance of the proposed method in distribution network reconfiguration

with optimal location and size of DG problems.

2015 Elsevier Ltd. All rights reserved.

Introduction

A distribution network is the last stage in delivery of electric

power. It carries electricity from the transmission system to con-

sumers. Distribution systems usually have high system losses

and poor voltage regulation because of the high current and low

voltage level in distribution systems [1,2]. In addition, due to the

rapid expansion of distribution networks, the voltage stability of

distribution systems has become an important issue. Therefore,

many efforts have been made to decrease the losses and improve

the voltage stability in distribution systems. Network reconfigura-

tion and distributed generator placement are among those effortsto mitigate this problem [3].

Distribution network reconfiguration (DNR) is the process of

varying the topology of distribution network by changing the

closed/open status of sectionalizing and tie switches while respect-

ing system constraints upon satisfying the operator’s objectives

[4]. The first publication about the DNR problem was presented

by Merlin and Back [5]. They solved DNR problem through a

discrete branch-and-bound type heuristic technique. Civanlar

et al. [6] proposed a switch exchange method to estimate the loss

reduction based on particular switching option. Since the method

is based on heuristics technique, it is difficult to take a systematic

way to evaluate an optimal solution. In recent years, new meta-

heuristic methods have been proposed for solving optimization

problems to obtain an optimal solution of global minimum in the

literature with good results. In [7], a method based on an enhanced

genetic algorithm was developed for DNR problem to minimize the

power loss and maximize the system reliability. Souza et al. [8]

proposed two new approaches for solving the DNR problem using

the Opt-aiNet (artificial immune network for optimization) andCopt-aiNet (artificial immune network for combinatorial optimiza-

tion) algorithms to minimize power loss. In [9], the network recon-

figuration and capacitor placement are simultaneously employed

to enhance the system efficiency in a fuzzy multi-objective opti-

mization problem by using a binary gravitational search algorithm

(BGSA).

Distributed generations (DGs), which are connected to the grid

at distributed level voltages are generating plant serving a cus-

tomer on-site. Because of the reasons of energy security and eco-

nomical benefit, the presence of DGs into distribution networks

has been increasing rapidly [10,11]. Impact of DG units on power

http://dx.doi.org/10.1016/j.ijepes.2015.12.030

0142-0615/ 2015 Elsevier Ltd. All rights reserved.

⇑ Corresponding author at: Dong An Polytechnic, 30/4 Str., Di An Dist., Binh

Duong Province, Viet Nam. Tel.: +84 0916664414.

E-mail address: [email protected] (T.T. Nguyen).

Electrical Power and Energy Systems 78 (2016) 801–815

Contents lists available at ScienceDirect

Electrical Power and Energy Systems

j o u r n a l h o m e p a g e : w w w . e l s e v i e r . c o m / l o c a t e / i j e p e s

http://dx.doi.org/10.1016/j.ijepes.2015.12.030mailto:[email protected]://dx.doi.org/10.1016/j.ijepes.2015.12.030http://www.sciencedirect.com/science/journal/01420615http://www.elsevier.com/locate/ijepeshttp://www.elsevier.com/locate/ijepeshttp://www.sciencedirect.com/science/journal/01420615http://dx.doi.org/10.1016/j.ijepes.2015.12.030mailto:[email protected]://dx.doi.org/10.1016/j.ijepes.2015.12.030http://-/?-http://-/?-http://-/?-http://-/?-http://-/?-http://-/?-http://-/?-http://crossmark.crossref.org/dialog/?doi=10.1016/j.ijepes.2015.12.030&domain=pdfhttp://-/?-

8/19/2019 1-s2.0-S0142061515005827-main.pdf

2/15

system has attracted the interest of some recent research efforts.

In [12], authors proposed comparison of novel power loss sensi-

tivity (NPLS), power stability index (PSI), and voltage stability

index (VSI) methods for optimal allocation and size of DG in

radial distribution network. In [13], a method based on bacterial

foraging optimization algorithm (BFOA) is proposed to find the

optimal location and size of DG with an objective of power losses

reduction, operational costs and improving voltage stability. In

[14], authors proposed a method based on the artificial neural

network to find optimal DG size and locations due to complexity

of multiple DG concepts. Kayal and Chanda [15] proposed a new

constrained multi-objective particle swarm optimization (PSO)

based wind turbine generation unit and photovoltaic array place-

ment approach to reduce power loss and improve voltage stabil-

ity of radial distribution system. In [16], a novel application of

multi-objective particle swarm optimization was developed for

determining the place and size of DGs, and the contract price of

their generated power.

Recently, some researches have integrated both the DNR and

DG placement problems to improve the effectively of distribution

network [17–19]. In [18], the DNR problem in the presence of DG

with an objective of minimizing real power loss and enhancing

voltage profile in distribution network is solved based on the har-

mony search algorithm (HSA). In [19], a method based on fireworksoptimization algorithm (FWA) is proposed for solving DNR

together with DG placement to minimize power loss and improve

voltage stability. Both researches [18,19] had used the different

techniques to pre-identify the candidate bus locations for DG

installation such as loss sensitivity factor (LSF) [18], and voltage

stability index (VSI) [19].

Due to pre-identifying of location of DGs based on LSF or VSI in

initial network configuration, the above methods focused only on

sizing of DG units. However, these parameters may change during

network reconfiguration process and DGs installation. In addition,

in distribution systems with multi-DG units, these parameters may

change more noticeable because of the interaction between DGs. In

this paper, a method based on cuckoo search algorithm (CSA) [20]

which is a recent meta-heuristic is proposed for solving the DNR problem in the presence of distributed generation. Compared to

other algorithms, CSA has fewer control parameters and is more

effective [4]. Recently, CSA has been applied to solve many power

system problems and other fields such as optimal power flow

(OPF) [21], power system stabilizers (PSSs) [22], load frequency

control (LFC) [23], and automatic generation control (AGC) [24].

The results obtained from the above problems have proven the

effectiveness of CSA compared to other optimization algorithms.

In this study, the proposed method based on CSA uses power

loss and VSI index as objective functions to find the optimum con-

figuration of distribution network, and the optimum bus location

and size of DGs. The algorithm is tested on 33-bus, 69-bus and

119-bus systems and results obtained are compared with other

techniques available in the literature.

Problem formulation

Objective functions

One of the main advantages of the optimal network reconfigu-

ration and DG installation is the reduction in power loss. The net

power loss reduced (DP Rloss) is taken as the ratio of total power loss

before and after the reconfiguration considering DGs of the system:

DP Rloss ¼ P rec :loss

P 0lossð1Þ

The total power loss of the system is determined by the summa-

tion of losses in all line sections:

P loss ¼XNbr i¼1

Ri P 2i þ Q

2

i

V 2i

! ð2Þ

On the other hand, as DGs are installed in distribution network,

the bus voltages will increase and voltage security will enhance.

Therefore, to obtain maximum benefit from the DG, suitable loca-

tion and sizing have to be determined before its installation based

on voltage stability index (VSI). VSI is a parameter that identifies

the near collapse nodes. The node with small VSI is more sensitive

to voltage collapse. According to Fig. 1, VSI of node 2th to node N this calculated as follows [25,16]:

VSIðkþ1Þ ¼ jV kj4 4 P kþ1 X k Q kþ1Rkð Þ

2

4 P kþ1Rk Q kþ1 X kð ÞjV kj2

ð3Þ

where P i+1, Q i+1 are total real power and reactive power load fed

through node (k + 1), respectively.

If the VSI for each bus is higher, the stability of that relevant

node shall be better. In distribution network reconfiguration con-

sidering DGs, the voltage stability deviation index (DVSI) can be

defined as follows:

DVSI ¼ max 1 VSIi

1 8i ¼ 2; . . . N bus ð4Þ

Sending node Receiving node

Pk+1 + jQk+1

R k +jXk

Vk

Vk+1

Ik k th

branch

Fig. 1. A branch of a radial distribution system.

Nomenclature

P rec loss total power loss of the system after reconfigurationP 0loss total power loss of the system before reconfigurationN br total number of branchesN bus total number of busesN DG number of distributed generators

P i real power load at bus iQ i reactive power load at bus iV i voltage magnitude at bus i

P Dgmax,i maximum size of distributed generator ithV min minimum acceptable bus voltageV max maximum acceptable bus voltageI i current at branch ithI max,i upper limit of line current as defined by the manufac-

turerRk resistance of the line section between buses k and k + 1 X k reactance of the line section between buses k and k + 1

802 T.T. Nguyen et al. / Electrical Power and Energy Systems 78 (2016) 801–815

http://-/?-http://-/?-http://-/?-http://-/?-http://-/?-http://-/?-http://-/?-

8/19/2019 1-s2.0-S0142061515005827-main.pdf

3/15

The proposed objective function (F ) of the problem is formu-

lated to minimize the total power loss and voltage stability

deviation index. The objective function can be described as

follows:

minimize F ¼ DP Rloss þ DVSI ð5Þ

Constraints

Each proposed configuration in DNR considering DGs process,

the power flow analysis should be carried out to calculate the volt-

age stability index, power loss of system and current of each

branch. The constraints of objective function are as follows:

(1) The computed voltages and currents should be in their

premising range for the proposed configuration.

V min 6 V i 6 V max; I ¼ 1; 2; . . . N bus ð6Þ

0 6 I i 6 I max;i; i ¼ 1;2; . . . N br ð7Þ

(2) The radial nature of distribution network must be main-

tained and all loads must be served.

(3) Distributed generation capacity limits:

0 6 P DGi 6 P DGmax;i; i ¼ 1;2; . . . N DG ð8Þ

Adaptive CSA for DNR considering DGs

CSA is a meta-heuristic search algorithm which has been pro-

posed recently by Yang and Deb [20]. The algorithm is inspired

based on the reproduction strategy of cuckoos. At the most basic

level, cuckoos lay their eggs in the nests of other host birds. The

host bird may discover that the eggs are not its own and either

destroy the egg or abandon the nest all together. To apply this as

an optimization tool, Yang and Deb [20] used three idealized rules:

At a time, each cuckoo lays one egg and puts its egg in a ran-domly chosen nest among the available host nests which is fixed.

The best nests with high quality of egg will be carried over to

the next generation.

A host bird can detect an alien egg. In this case, it can either

throw the egg away or abandon the nest so as to build a new nest

in a new location.

When the CSA algorithm is employed to solve the DNR consid-

ering DGs. The radial constraint imposes the main hurdle since a

large number of infeasible individuals appear during initial stage

and intermediate stages of the evolutionary process because a set

of definite number of tie lines will act as a host nests, such that

when they open, a feasible radial configuration is formed or not.

Therefore, the CSA needs modification using some engineering

knowledge base to make it compatible with the DNR problem. Inthe proposed adaptive CSA (ACSA), the host nests generating are

adapted using graph theory to reduce the number of infeasible

individuals at each stage of the optimization process.

Variable expression for DNR

In the conventional CSA, the initial population is randomly cre-

ated, which consists of a large number of infeasible individuals vio-

lating the radial constraint, particularly in medium/large

distribution networks. In the proposed ACSA, these infeasible indi-

viduals are reduced by using graph theory. A distribution network

can be represented with a graph that contains a set of branches B

and a set of nodes N . They are presented by connection matrix A,

which has one row for each branch and one column for each node.In this matrix if any branch i is directed away from node j, element

is equal 1, and if any branch i is directed toward node j, element is

equal 1, otherwise its element is zero [26].

Fundamental loops (FLs) based on a graph theory are deter-

mined for the meshed network by closing all tie-switches. It is

found that, the number of fundamental loops of the system is

equal to the number of normal open branches (NO) [27,28]. For

finding FLs of the network after determining connection matrix

A, in each level, a normal open branch is added to system to form

a new loop [28]. Based on the method proposed in [28], branches

connected to the nodes, which have sum of absolute each ele-

ment of a corresponding column in the connection matrix A is

1, are removed. This process is repeated until this node type no

longer remains in the system. Finally the numbers of remaining

branches can be saved in a vector that this vector represents a

FL [28]. Fig. 2 shows the topology of a 16-node system and the

first FL is determined in Fig. 3. As presented in Fig. 3, after deter-

mining connection matrix A, to identify the first FL, the following

steps are done.

Step 1: Add branch 14 to the connection matrix A.

Step 2: Calculate sum of absolute each element of a correspond-

ing column in matrix A, the nodes which have the result is 1, are

node 3, 8, 9, 10, and 12. Therefore, branch 2, 6, 12, 13, and 8 are

removed. Because these branches connected to node 3, 8, 9, 10,

and 12 respectively.

Step 3: Similarly to step 2, calculate sum of absolute each ele-

ment of a corresponding column for rest branches in matrix A

and the result is the branch 5 removed.

Step 4: Similarly to step 2, calculate sum of absolute each ele-

ment of a corresponding column for rest branches in matrix A

and the result is the branch 4 removed.

Step 5: There are not any nodes that have sum of absolute each

element of a corresponding column in matrix A are 1. Therefore,

the first FL consists of branches {1, 3, 7, 9, 10, 11, and 14}. Sim-

ilarly to branch 14, by adding branch 15 the second FL which

consists of branches {1, 2, 4, 5, 12, and 15} will be obtained.

Continually, the third FL with branches {4, 6, 7, 8, and 16} willbe obtained by adding branch to matrix A. Pseudo code of the

finding fundamental loop algorithms is given in Fig. 4.

Each radial configuration which involves a set of open branches

are randomly chosen from corresponding FLs. This helps to reduce

the generation of infeasible configuration during each stage of the

optimization algorithm. However, some of the branches are com-

mon in some of FLs [28]. Therefore, radial condition of network

must be checked. In each configuration, the connection matrix A

is determined. Then, the first column corresponding to the refer-

ence node in the network will be removed to form a square matrix

if the configuration of the network is radial. It is found that, if the

1 1 1

4 5 1314

2

3 9

7

6

8 12

11

10

s1

s7 s4

s3

s14 s11 s10

s9

s8

s16

s6 s5

s13

s12

s15

s2

Fig. 2. The topology of a 16-node system.

T.T. Nguyen et al./ Electrical Power and Energy Systems 78 (2016) 801–815 803

http://-/?-http://-/?-http://-/?-http://-/?-http://-/?-

8/19/2019 1-s2.0-S0142061515005827-main.pdf

4/15

determinant of square matrix A is equal to 1 or1 then the system

is radial [26]. Pseudo code of the checking system radially algo-

rithm is given in Fig. 5.

Implementation of adaptive CSA for DNR considering DGs

Based on the three rules of CSA in Section ‘Adaptive CSA for DNR

considering DGs’, the ACSA method is implemented for DNR con-

sidering DGs as follows:

Step 1: Determine a fundamental loops.

Step 2: Determine the upper bound and lower bound of each

tie-line based on the size of the corresponding fundamental

loop.

Step 3: Initialization.

In DNR process using ACSA, each radial structure of the network

is represented as a host nest. A population of N host nests is repre-

sented by X i ¼ X i1; . . . ; X

id1; X

id

h i with i = 1, 2 . . . N . To solve DNR

problemwith simultaneous DG allocation, the first part of the solu-

tion vector is takenas the number of open branches in the distribu-

tion network, and the second part is the number of buseschosen for

DG installation and the third part is the sizes of DGs. Thus the solu-

tion vector for simultaneous reconfiguration and DG installation

considering their location and size is formed as follows:

X i ¼ Tiei1; . . . ;Tie

iNO; Lo:DG

i1; . . . ; Lo:DG

im; Size:DG

i1; . . . ; Size:DG

im

h ið9Þ

where Tie1, Tie2, . . . , TieNO are open branches in the fundamental

loops FL1 to FLNO formed corresponding to the tie lines; Lo.DG1, Lo.

DG2, . . . , Lo.DGm are the buses for DG installation; and Size.DG1,

Size.DG2, . . . , Size.DGm are sizes of DG units (MW) to be installed

at buses respectively. NO and m are the number of tie lines and

the number of DGs, respectively.

In the ACSA, each egg can be regarded as a solution which is

randomly generated in the initialization. Therefore, each nest i of the population is randomly initialized as follows:

Fig. 3. Determination FL when closing branch 14.

Input: Connected matrix A for the initial configuration of distribution network,initial open branches.

Output: Fundamental loopsFor (k: =1 to the number of normal open branches) do

Add the normal open branch k to the matrix ASum_column:= Sum of absolute each element of each column in the

matrix A

While (Sum_column jth =1, j=1… N) doFor (i=1 to the number of branches) do

If element (i, j) =1 or -1 then Remove the branch ith from matrix A

End if

End for iUpdate sum of absolute elements of each column in the matrix A

End whileSave the remaining branches of matrix A to the FL k

th

End for k

Fig. 4. Pseudo code of the finding fundamental loop algorithms.


8/19/2019 1-s2.0-S0142061515005827-main.pdf

5/15

Tiei ¼ round Tieilower ;d1 þ rand ðTie

iupper ;d1 Tie

ilower ;d1Þ

h i ð10Þ

Lo:DGi ¼ round Loilower ;d2 þ rand ðLo

iupper ;d2 Lo

ilower ;d2Þ

h i ð11Þ

Size:DGi ¼ Sizeilower ;d3 þ rand ðSize

iupper ;d3 Size

ilower ;d3Þ

h i ð12Þ

where d1 = 1, 2 . . . NO, d2 = 1, 2 . . . m and d3 = 1, 2 . . . m. Tielower ,d1and Tie upper ,d1 are minimum tie-line and maximum tie-line which

are encoded in fundamental loop d1. DGs are located from any

nodes excluding the reference node. Therefore, the lower bound

(Lolower ,d2) and upper bound (Loupper ,d2) of each location of DG is

from node 2 to the maximum node of the network and the size of

each DG is from zero to maximum power of DG.

Based on the initialized population of the nests, the radial topol-

ogy checking algorithm is run to check the nests and the objectivefunction of each nest is calculated by the power flows using New-

ton–Raphson method. Based on the values of objective function,

the nest with the best fitness function is saved to the best nest

Gbest .

Step 4: Generation of new solution via Lévy flight

Except for the best nest, all the other nests are replaced based

on the quality of new cuckoo eggs which are generated by Lévy

flights from their position as follows:

X newd ¼ Xbest d þ a rand D X newd ð13Þ

where a > 0 is the step size parameter; rand is a random number in

range [0,1] and the increased value D X newi is determined by:

D X newi ¼ rand x

jrand yj1=b

r xðbÞr yðbÞ

ð Xbest i Gbest iÞ ð14Þ

where rand x and rand y are two normally distributed stochastic vari-

ables with standard deviation r xðbÞ and r yðbÞ given by:

r xðbÞ ¼ Cð1 þ bÞ sinðpb

2 Þ

Cð1þb2

Þ b 2ðb1

2 Þ

24

35

1=b

ð15Þ

r yðbÞ ¼ 1 ð16Þ

where b is the distribution factor ð0:

36 b 6 1:

99Þ and C is thegamma distribution function.

Some nests of population may violate the boundary condition of

the optimization problem. Therefore, they are redefined at the end

of the generation of new solution process, as follows:

X newd ¼ ½round½ X newd1 ; round½ X

newd2 ; X

newd3 ð17Þ

The limits for each tie-line, location of each DG and power are

determined as follows:

Tielimd1 ¼

Tielower ;d1 if Tied1 < Tielower ;d1

Tieupper ;d1 if Tied1 > Tieupper ;d1

Tied1 otherwise

8>>><>>>:

ð18Þ

Lo:DGlimd2 ¼

2 if Lo:DGd1 < 2

Loupper ;d2 if Lod2 > Loupper ;d2

Lo:DGd2 otherwise

8>>><>>>:

ð19Þ

Size:DGlimd3 ¼

Sizelower ;d3 if Sized3 < Sizelower ;d3

Sizeupper ;d3 if Sized3 > Sizeupper ;d3

Size:DGd3 otherwise

8><>: ð20Þ

Based on the new population of the nests, the radial topology

checking algorithm is run to check the nests. Then, the fitness val-

ues are calculated to find the best value of each nest Xbest d.

Step 5: Alien eggs discovery

A fraction pa of the worst nests can be abandoned so that new

nests can be built at new locations by random walks. Existing eggs

will be replaced by a good quality of new generated ones from

their current positions through random walks with step size as

follows:

X newi ¼ Xbest i þ K D X newi ð21Þ

where K is the updated coefficient determined based on the proba-

bility of a host bird to discover an alien egg in its nest:

K ¼ 1 if rand < P a

0; otherwise

ð22Þ

And the increased value D X newi is determined by:

D X newi ¼ rand randp1ð Xbest iÞ randp2ð Xbest iÞ½ ð23Þ

Input: a candidate configuration with set of open branchesOutput: a candidate configuration is a radial configuration or not

Determine connection matrix A for the network, which involves initial openbranches.

Remove the first column of matrix A Remove the rows of matrix A corresponding to open branches in candidate

configuration

If (matrix A is a square matrix)Calculate determinant of square matrix A

If (determinant of square matrix A = 1 or -1)Output: = a candidate configuration is a radial configuration

Else

Output: = a candidate configuration is not a radial configuration End if

Else

Output: = a candidate configuration is not a radial configuration

End if

Fig. 5. Pseudo code of the checking system radially algorithm.


http://-/?-http://-/?-http://-/?-http://-/?-http://-/?-http://-/?-http://-/?-http://-/?-http://-/?-http://-/?-

8/19/2019 1-s2.0-S0142061515005827-main.pdf

6/15

where rand is the random numbers in [0, 1] and

randp1ð Xbest iÞ; randp2ð Xbest iÞ are therandom perturbation for posi-tions of the nests in Xbest i: Some nests of population may violate

the boundary condition of the optimization problem. Hence, they

are redefined at the end of the generation of new solution process,

as Eq. (17):

X newd ¼ ½round½ X newd1 ; round½ X

newd2 ; X

newd3 ð17Þ

For the new solution, its lower and upper limits should be sat-

isfied according to their limits by using Eqs. (18)–(20).

Based on the new population of the nests, the radial topology

checking algorithm is run to check the nests. Then, the fitness val-

ues are calculated by the power flows using Newton–Raphson

method to find the best value of each nest Xbest d and the nest cor-responding to the best fitness value is set to the best nest Gbest .

Step 6: Termination criterion

The generating new cuckoos by Lévy flight and discovering alien

eggs steps are alternatively performed until the number of itera-

tions (Iter ) reaches the maximum number of iterations (Iter max).

The Pseudo code of the proposed ACSA for DNR considering DGs

is given in Fig. 6.

Application results and analysis

In order to demonstrate and examine the applicability of the

proposed technique in solving the network reconfiguration and

installation of DG units simultaneously from small scale to large

scale distribution networks using ACSA, it is applied to three test

systems consisting of 33 buses, 69 buses and 119 buses. The max-imum number of DGs installed for the given test systems is limited

Input: line and load data of the distribution network

Output: optimal configuration and optimal the location and size of DGs

Step 1: Determine a fundamental loops.

Step 2: Determine the upper bound and lower bound of each tie-line based on size

of the corresponding fundamental loops.

Step 3: Generate initial population of N host nests with i =

1, 2… N .Check radially condition of each host nests by checking system radially

algorithm.

If hosts nest Xi is radial configuration then

Calculate the fitness function of Xi to find the best nest Gbest.

Else

Fitness function of Xi = inf

End if

Step 6: While (Iter < Iter max) do

Step 4: Get new solution by Lévy flight

Redefine the new solution depending on the boundary condition

Check radially condition of each host nests by checking system radially

algorithm


Evaluate fitness function to choose new Xbest d

Else


End if

Step 5: Get new solution by Alien eggs discovery

Redefine the new solution depending on the boundary condition

Check radially condition of each host nests by checking system radially

algorithm


Evaluate fitness function to choose new Xbest d and Gbest

Else


End if

If fitness (Gbest) < Fmin then

Fmin = fitness (Gbest)

Best nest = Gbest

End if

End While

Post process result: best fitness value Fmin and the Best nest

Fig. 6. Pseudo code of the ACSA for DNR considering DGs algorithm.


8/19/2019 1-s2.0-S0142061515005827-main.pdf

7/15

to three. The limits of DG unit sizes chosen for installation are 0 to

2 MW for 33-bus and 69-bus system and 0 to 5 MW for 119-bus

system. In the simulation of network, seven scenarios are consid-

ered to analyze the superiority of the proposed method.

Scenario 1: Base case (without reconfiguration and distributed

generators).

Scenario 2: The system is only reconfigured.

Scenario 3: Allocation and size of DGs are optimized on base

case.

Scenario 4: Allocation and size of DGs are optimized after

reconfiguration of the network.

Scenario 5: The system is reconfigured after DGs installed based

on scenario 3.

Scenario 6: The System is simultaneous reconfigured and opti-

mized size of DGs (VSI for all nodes of the system is computed

from power flow to identify the candidate bus location of DGs).

Scenario 7: The System is simultaneous reconfigured and opti-

mized allocation and size of DGs.

33-bus test system

The 33-bus distribution system, which is a small-scale distribu-

tion networks, includes 37 branches, 32 sectionalizing switches

and 5 tie switches. The line and load data of this system are taken

from [29]. The total real and reactive power loads of the system are

3.72 MW and 2.3 MVAr, respectively. Fig. 7 shows the single line

diagram of this network. The parameters of ACSA algorithm used

in the simulation of network are number of nets N p = 30, probabil-

ity of an alien egg to be discovered Pa = 0.2, number of iterations

Iter max = 2000.

Fundamental loops of the network are obtained by the finding

fundamental loop algorithms as given in Table 1. The performance

of the proposed method is presented in Table 2. From Table 2, it

can be seen clearly that in the initial case, power loss (kW) in the

system is 202.68, which is reduced to 139.98, 74.26, 58.79, 62.98,63.69, and 53.21 using scenarios 2, 3, 4, 5, 6, and 7, respectively.

The percentage power loss reduction for scenario 2–7 is 30.93,

63.26, 71.0, 68.93, 68.58, and 73.75, respectively. It can be also

seen from Table 2 that, the minimum voltage magnitude of the sys-

tem is improved remarkably in all the scenarios. In the base case,

the minimum voltage magnitude is improved from 0.9108 p.u. to

0.9413, 0.9778, 0.9802, 0.9826, 0.9786, and 0.9806 p.u. for using

case 2 to case 7. In addition, the VSI is also improved from

0.6978 to 0.7878, 0.9118, 0.9264, 0.9354, 0.9202, and 0.9318 by

using scenarios 2, 3, 4, 5, 6 and 7, respectively. It is observed that

the power loss reduced using scenario 7 is the highest, which

demonstrates that the bus location of DGs need to optimize simul-

taneously with the reconfiguration and optimum size of DGs pro-

cess. The voltage profiles (which are shown in node voltages and

VSI of nodes) of all seven scenarios are compared and shown in

Figs. 8 and 9. From the figures, it is observed that the voltage pro-

file at all buses has been improved significantly after using recon-

figuration and optimization of location and size of DGs.

In order to illustrate the performance of the proposed method,

the performance of ACSA is compared with the results of fireworks

algorithm (FWA) [19] and harmony search algorithm (HSA) [18]

available in the literature and is presented in Table 2. From the

table, it is perceived that at all scenarios, the performance of the

ACSA is better than FWA and HSA in terms of power loss and min-

imum voltage. The convergence results of system indices from sce-

nario 2 to scenario 6 are shown in Fig. 10. It can be seen from

Fig. 10 that the fitness value of scenario 7 is the lowest compared

to other scenarios.

69-bus test system

To demonstrate the applicability of the proposed method using

ACSA in medium-scale system. It is simulated on 69-bus system.

The 69-bus distribution system includes 69 nodes, 73 branches.There are 5 tie switches and total loads are 3.802 MW and

2.696 MVAr [30]. The schematic diagram of the test system is

shown in Fig. 11. In a normal operation, switches {69, 70, 71, 72,

and 73} are opened. The parameters of ACSA algorithm used in

the simulation of network are number of nets N p = 30, probability

of an alien egg to be discovered Pa = 0.2, number of iterations

Iter max = 2000.

Similar to 33-bus test system, fundamental loops of the net-

work are also obtained by the finding fundamental loop algorithms

5 46 82 3 7

19

9 11 21 31 41 1615 1817

26 27 28 29 30 31 32 33

23 24 25

20 21 22

10

2 3 54 6 7

18

19 20

33

1 9 10 11 12 13 14

34

8

21 35

15 16 17

25

26 27 28 29 30 31 32 36

37

22

23 24

1

Fig. 7. IEEE 33-bus test system.

Table 1

Fundamental loops of the 33-bus system.

Fundamental

loop

Tie-line

FL 1 2, 3, 4, 5, 6, 7, 18, 19, 20, 33

FL 2 9, 10, 11, 12, 13, 14, 34

FL 3 2, 3, 4, 5, 6, 7, 8, 9, 10, 11, 18, 19, 20, 21, 35

FL 4 6, 7, 8, 9, 10, 11, 12, 13, 14, 15, 16, 17, 25, 26, 27, 28, 29, 30,

31, 32, 36FL 5 3, 4, 5, 22, 23, 24, 25, 26, 27, 28, 37


http://-/?-http://-/?-

8/19/2019 1-s2.0-S0142061515005827-main.pdf

8/15

Table 2

Performance analysis of proposed method on the 33-bus system.

Scenario Item Proposed ACSA FWA [19] HSA [18]

Base case (Scenario 1) Switches opened 33, 34, 35, 36, 37 – –

Power loss (kW) 202.68 – –

Minimum voltage (p.u.) 0.9108 – –

Minimum VSI 0.6978 – –

Only reconfiguration (Scenario 2) Switches opened 07, 14, 9, 32, 28 7, 14, 9, 32, 28 7, 14, 9, 32, 37

Power loss (kW) 139.98 139.98 138.06

% Loss reduction 30.93 30.93 31.88

Minimum voltage (p.u.) 0.9413 0.9413 0.9342


Only DG installation (Scenario 3) Switches opened 33, 34, 35, 36, 37 33, 34, 35, 36, 37 33, 34, 35, 36, 37

Size of DG in MW (Bus number) 0.7798 (14) 0.5897 (14) 0.1070 (18)

1.1251 (24) 0.1895 (18) 0.5724 (17)

1.3496 (30) 1.0146 (32) 1.0462 (33)

Power loss (kW) 74.26 88.68 96.76

% Loss reduction 63.26 56.24 52.26



DG installation after reconfiguration

(Scenario 4)

Switches opened 7, 14, 9, 32, 28 7, 14, 9, 32, 28 7, 14, 9, 32, 37


0.5397 (12) 0.3141 (33) 0.1611 (31)

0.5045 (16) 0.1591 (18) 0.6612 (30)

Power loss (kW) 58.79 83.91 97.13% Loss reduction 71.00 58.59 52.07



Reconfiguration after DG installation

(Scenario 5)

Switches opened 33, 9, 8, 36, 27 7, 34, 9, 32, 28 –

Size of DG in MW (Bus number) 0.7798 (14) 0.5897 (14) –

1.1251 (24) 0.1895 (18)

1.3496 (30) 1.0146 (32)

Power loss (kW) 62.98 68.28 –

% Loss reduction 68.93 66.31 –

Minimum voltage (p.u.) 0.9826 0.9712 –


Simultaneous Reconfiguration and DG

installation (Scenario 6)

Switches opened 7, 10, 13, 32, 27 7, 14, 11, 32, 28 7, 14, 10, 32, 28


1.2024 (29) 0.6158 (29) 0.5586 (31)

0.7127 (18) 0.5315 (18) 0.5840 (33)

Power loss (kW) 63.69 67.11 73.05

% Loss reduction 68.58 66.89 63.95



Simultaneous Reconfiguration, DG

installation and location of DG

(Scenario 7)

Switches opened 33, 34, 11, 31, 28 – –

Size of DG in MW (Bus number) 0.8968 (18) – –

1.4381 (25)

0.9646 (7)


% Loss reduction 73.75 – –



5 10 15 20 25 30 330.91

0.92

0.93

0.94

0.95

0.96

0.97

0.98

0.99

1

Node No.

V o l t a g e

( p . u . )

Case 1

Case 2

Case 3

Case 4

Case 5

Case 6

Case 7

Fig. 8. Comparison of node voltages of 33-bus system.


8/19/2019 1-s2.0-S0142061515005827-main.pdf

9/15

0 5 10 15 20 25 30 33

0.7

0.75

0.8

0.85

0.9

0.95

1

Node No.

V

S I

Case 1

Case 2

Case 3

Case 4

Case 5

Case 6

Case 7

Fig. 9. Comparison of VSI-nodes of 33-bus system.

0 200 400 600 800 1000 1200 1400 1600 1800 20000.3

0.4

0.5

0.6

0.7

0.8

0.9

1

Iterations

F i t n e s s

v a l u e

Case 2: Only Rec.

Case 3: Only DG ins.

Case 4: DG ins. after Rec.

Case 5: Rec. after DG ins.

Case 6: Rec. and size DG ins.

Case 7: Rec., loc. and siz. DG

Fig. 10. Comparison of 33-bus system indices for scenarios in fitness function.

154 6 82

3

7 1911 219 1413 1615 1817 27

66 67

23 24 2520 21 22 26

68 69

10

36 37 83 93 544424 341404 46

586545 5535 596560 16 26

6463

47 4948 50

3328 3029 13 23 34 35

1 2 54 6 7 8 9 1211 14 1310 1951 6 1 1817

27

23 24 2520 21 22 26

3328 3029 3231 34

35

37 3938 40 41 4342 44 45

46

36

84 94 7 4

515250

51

52

575853 4 5 55 56 59

65

60 16 26 36 4 6 57

66

67

68

69

70

71

72

73

Fig. 11. IEEE 69-bus test system.


http://-/?-http://-/?-http://-/?-

8/19/2019 1-s2.0-S0142061515005827-main.pdf

10/15

as given in Table 3. This test system is also simulated for seven sce-

narios and the results are presented in Table 4. It is observed from

Table 4, base case power loss (in kW) in the systemis 224.89 which

is reduced to 98.59, 72.44, 37.23, 41.13, 40.49, and 37.02 using sce-

narios 2, 3, 4, 5, 6, and 7, respectively. The percentage loss reduc-

tion for scenario 2–7 is 56.16, 67.79, 83.45, 81.71, 82.0, and

83.54, respectively. In six scenarios, power loss reduction using

scenario 7, which is the proposed method, is the highest, which

elicits the superiority of the proposed method over the others.

From Table 4, it is also seen that improvement in power loss reduc-

tion and voltage profile for scenario 7 are higher when compared to

other scenarios. This illustrates that reconfiguration presence DGs

need to concern simultaneously the location and size of DG.

By using scenario 2 to scenario 7, the minimum voltage magni-

tude is improved from 0.9092 p.u. to 0.9495, 0.9890, 0.9870,

0.9828, 0.9873, and 0.9869 p.u. The VSI is enhanced significantly

from 0.6859 to 0.8414, 0.9546, 0.9390, 0.9260, 0.9403, and

0.9433. The voltage profiles of the system for seven scenarios are

compared and shown in Figs. 12 and 13. From the figures, it is

observed that the voltage profile at all buses has been improved

significantly after using proposed method. Fig. 14 shows the con-

vergence characteristics of the algorithms for the best solution in

Table 3

Fundamental loops of the 69-bus system.

Fundamental

loop

Tie-line

FL 1 3, 4, 5, 6, 7, 8, 9, 10, 35, 36, 37, 38, 39, 40, 41, 42, 69

FL 2 13, 14, 15, 16, 17, 18, 19, 20, 70

FL 3 3, 4, 5, 6, 7, 8, 9, 10, 11, 12, 13, 14, 35, 36, 37, 38, 39, 40, 41,

42, 43, 44, 45, 71

FL 4 4, 5, 6, 7, 8, 46, 47, 48, 49, 52, 53, 54, 55, 56, 57, 58, 72FL 5 9, 10, 11, 12, 13, 14, 15, 16, 17, 18, 19, 20, 21, 22, 23, 24, 25,

26, 52, 53, 54, 55, 56, 57 58, 59, 60, 61, 62, 63, 64, 73

Table 4


Scenario Item Proposed ACSA FWA [19] HSA [18]Base case (Scenario 1) Switches opened 69, 70, 71, 72, 73 – –




Only reconfiguration (Scenario

2)

Switches opened 69, 70, 14, 57, 61 69, 70, 14, 56, 61 69, 18, 13, 56, 61

Power loss (kW) 98.59 98.59 99.35

% Loss reduction 56.16 56.16 55.85



Only DG installation (Scenario 3) Switches opened 69, 70, 71, 72, 73 69, 70, 71, 72, 73


0.3804 (18) 1.1986 (61) 0.3690 (64)

2 (61) 0.2258 (27) 1.3024 (63)

Power loss (kW) 72.44 77.85 86.77

% Loss reduction 67.79 65.39 61.43

Minimum voltage (p.u.) 0.9890 0.9740 0.9677Minimum VSI 0.9546 – –

DG installation after

reconfiguration (Scenario 4)

Switches opened 69, 70, 14, 57, 61 69, 70, 14, 56, 61 69, 18, 13, 56, 61


0.4666 (64) 0.2145 (62) 0.3525 (60)

0.3686 (12) 0.1425 (64) 0.4257 (58)

Power loss (kW) 37.23 43.88 51.3

% Loss reduction 83.45 80.49 77.2



Reconfiguration after DG

installation (Scenario 5)

Switches opened 69, 70, 14, 58, 64 69, 70, 12, 58, 61 –

Size of DG in MW (Bus number) 0.6022 (11) 0.4085 (65) –

0.3804 (18) 1.1986 (61)

2 (61) 0.2258 (27)

Power loss (kW) 41.13 39.69 –

% Loss reduction 81.71 82.35 –

Minimum voltage (p.u.) 0.9828 0.9763 –


Simultaneous Reconfiguration

and DG installation (Scenario

6)

Switches opened 69, 70, 12, 58, 61 69, 70, 13, 55, 63 69, 17, 13, 58, 61


0.1566 (62) 0.2750 (62) 0.3525 (60)

0.4090 (65) 0.4159 (65) 0.4257 (62)

Power loss (kW) 40.49 39.25 40.3

% Loss reduction 82.0 82.55 82.08



Simultaneous Reconfiguration,

DG installation and location

of DG (Scenario 7)

Switches opened 69, 70, 14, 58, 61 – –

Size of DG in MW (Bus number) 0.5413 (11) – –

0.5536 (65)

1.7240 (61)


% Loss reduction 83.54 – –




8/19/2019 1-s2.0-S0142061515005827-main.pdf

11/15

six scenarios. From the figure, the fitness function in scenario 7 is

the most minimum in six scenarios.

Similar to 33-bus test system, the performance of ACSA on

69-bus system is also compared with the results of FWA and

HSA, and the results are presented in Table 4. From the table, it

is observed that the performance of the ACSA is better compared

to HSA in terms of the quality of solutions in most scenarios. Also

from Table 4, the best results are identical to results obtained by

FWA in the scenario 2. It is also shown from Table 4 that the ACSA

has outperformed FWA in terms of power loss minimization and

voltage stability enhancement in the scenario 3 and 4. In the sce-

nario 5 and 6, although the proposed algorithm finds the optimal

configuration with the power loss reduction in percent are 81.71

and 82.0, respectively. This values are 0.64 and 0.55 lower than

10 20 30 40 50 60 69

0.91

0.92

0.93

0.94

0.95

0.96

0.97

0.98

0.99

1

Node No.

V o l t a

g e

( p . u . )

Case 1

Case 2

Case 3

Case 4

Case 5

Case 6

Case 7


0 10 20 30 40 50 60 69

0.7

0.75

0.8

0.85

0.9

0.95

1

Node No.

V S I

Case 1

Case 2

Case 3

Case 4

Case 5

Case 6

Case 7


0 200 400 600 800 1000 1200 1400 1600 1800 2000

0.2

0.3

0.4

0.5

0.6

0.7

0.8

0.9

1

Iterations

F i t n e s s

v a l u e

Case 2: Only Rec.






Fig. 14. Comparison of 69-bus system indices for scenarios in fitness function.


http://-/?-http://-/?-http://-/?-

8/19/2019 1-s2.0-S0142061515005827-main.pdf

12/15

8/19/2019 1-s2.0-S0142061515005827-main.pdf

13/15

selected for scenario 6 and 7 are Iter max = 5000 and Iter max = 2000

for the rest of scenarios.

The fundamental loops of the network are obtained by the

finding fundamental loop algorithms as given in Table 5. Table 6show the results obtained from proposed method for seven

scenarios. It can be seen from Table 6, base case power loss (in

kW) in the system is 1273.45 which is reduced to 855.04,

648.10, 631.19, 613.79, 682.09, and 586.24 using scenarios 2, 3,

4, 5, 6, and 7, respectively. The percentage loss reduction forscenario 2–7 is 32.86, 49.11, 50.43, 51.80, 46.44, and 53.96,

Table 6


Scenario Item Proposed ACSA

Base case (Scenario 1) Switches opened 118, 119, 120, 121, 122, 123, 124, 125, 126, 127, 128, 129, 130,

131, 132

Power loss (kW) 1273.45

Minimum voltage (p.u.) 0.8678

Minimum VSI 0.5676

Only reconfiguration (Scenario 2) Switches opened 42, 25, 23, 121, 50, 58, 39, 95, 71, 74, 97, 129, 130, 109, 34


% Loss reduction 32.86


Minimum VSI 0.7535

Only DG installation (Scenario 3) Switches opened 118, 119, 120, 121, 122, 123, 124, 125, 126, 127, 128, 129, 130,

131, 132

Size of DG in MW (Bus

number)

3.2664 (71)

3.1203 (109)

2.86267 (50)




Minimum VSI 0.8199

DG installation after reconfiguration (Scenario 4) Switches opened 42, 25, 23, 121, 50, 58, 39, 95, 71, 74, 97, 129, 130, 109, 34


number)

1.7145 (111)

1.7565 (96)5 (65)




Minimum VSI 0.8208

Reconfiguration after DG installation (Scenario 5) Switches opened 42, 25, 21, 121, 122, 58, 39, 125, 70, 127, 128, 129, 85, 131, 33


number)

3.2664 (71)

3.1203 (109)

2.86267 (50)




Minimum VSI 0.8523

Simultaneous Reconfiguration and DG installation (Scenario 6) Switches opened 42, 25, 23, 121, 50, 61, 39, 125, 126, 70, 75, 129, 130, 109, 34


number)

2.9585 (75)

0.1924 (76)

1.3397 (77)




Minimum VSI 0.7535

Simultaneous Reconfiguration, DG installation and location of DG

(Scenario 7)

Switches opened 42, 25, 22, 121, 122, 58, 39, 125, 70, 127, 128, 81, 130, 131, 33


number)

2.5331 (50)

3.6819 (109)

3.7043 (73)




Minimum VSI 0.8700

Table 7

Comparison of simulation results for 119-node network in Scenario 2.

Methods Open Switches Delta P (kW) V min (p.u.)

Initial 119, 120, 121, 122, 123, 124, 125, 126, 127, 128, 129, 130, 131, 132, 133 1273.45 0.8678

Proposed ACSA 42, 25, 23, 121, 50, 58, 39, 95, 71, 74, 97, 129, 130, 109, 34 855.04 0.9298

ITS [31] 42, 26, 23, 51, 122, 58, 39, 95, 71, 74, 97, 129, 130, 109, 34 867.4 0.9323

MTS [26] 42, 26, 23, 51, 122, 58, 39, 95, 71, 74, 97, 129, 130, 109, 34 867.4 0.9323

CSA [4] 42, 25, 23, 121, 50, 58, 39, 95, 71, 74, 97, 129, 130, 109, 34 855.04 0.9298

FWA [32] 42, 25, 23, 121, 50, 58, 39, 95, 71, 74, 97, 129, 130, 109, 34 855.04 0.9298


http://-/?-http://-/?-

8/19/2019 1-s2.0-S0142061515005827-main.pdf

14/15

respectively. In six scenarios, power loss reduction using scenario

7, which is the proposed method, is the highest, which elicits the

superiority of the proposed method over the others. By using case

2 to case 7, the minimum voltage magnitude is improved from

0.8678 p.u. to 0.9298, 0.9515, 0.9526, 0.9608, 0.9298, and

0.9644 p.u. Similar to the minimum voltage magnitude, the VSI

is also improved from 0.5676 to 0.7535, 0.8199, 0.8208, 0.8523,

0.7535, and 0.87 by using scenarios 2, 3, 4, 5, 6 and 7, respec-

tively. From Table 6, it is seen that enhancement in power loss

reduction and voltage profile for scenario 7 are higher when com-

pared to scenario 6. This implies that simultaneous reconfigured

and optimized size of DGs, which is the scenario 6, does not yield

0 20 40 60 80 100 118

0.88

0.9

0.92

0.94

0.96

0.98

1

Node No.

V o l t a g

e ( p . u . )

Case 1

Case 2

Case 3

Case 4

Case 5

Case 6

Case 7


0 20 40 60 80 100 1180.55

0.6

0.65

0.7

0.75

0.8

0.85

0.9

0.95

1

Node No.

V S I

Case 1

Case 2

Case 3

Case 4

Case 5

Case 6

Case 7


0 500 1000 1500 2000 2500 3000 3500 4000 4500 5000

0.6

0.8

1

1.2

1.4

1.6

1.8

2

Iteration

F i t n e s s V a l u e

Case 2: Only Rec.






Fig. 18. Comparison of performance of ACSA in six scenarios for minimization of the 119-bus system.


8/19/2019 1-s2.0-S0142061515005827-main.pdf

15/15

desired results of minimizing power loss and maximizing voltage

profile.

Table 7 presents the comparison is made with previous study in

scenario 2. These results are identical to the results obtained by the

methods proposed in Refs. [4,32] and better than the results

obtained by the modified tabu search (MTS) algorithm [26] and

the improved tabu search (ITS) algorithm [31]. The voltage profiles

of the system for seven scenarios are compared and shown in

Figs. 16 and 17. From the figures, it is observed that the voltage

profile has been improved drastically after using proposed method.

Fig. 18 shows the convergence characteristics of the algorithms for

the best solution in six scenarios. As illustrated in this figure, the

fitness function in scenario 7 is the most minimum in six scenarios.

This results elicit the superiority of the proposed method.

Conclusion

In this paper, the ACSA method has been successfully applied

for distribution network reconfiguration and simultaneous loca-

tion and size of DG problem. The objective is to minimize the active

power loss and enhance voltage stability index of power distribu-

tion systems. In addition, different loss reduction and enhance-

ment voltage stability methods such as only network

reconfiguration, only DG installation, DG installation after recon-

figuration, reconfiguration after DG installation, simultaneous

reconfiguration and DG installation are also simulated to establish

the superiority of the proposed method. The proposed method

based on graph theory is used to determine the search space of

each tie-line, which helps the cuckoo search algorithm reduces

infeasible network configurations at each stage of the optimization

process and it is also is adapted to check the radial constraint of

each generated configuration. The proposed method is tested on

33-bus, 69-bus, and 119-bus test systems. The results demonstrate

that network reconfiguration with simultaneous location and size

of DG is more effective in reducing power loss and improving the

voltage profile compared to other scenarios. The simulated results

are also compared with the results of FWA and HSA available in the

literature. The computational results have demonstrated that the

performance of the ACSA is better than FWA and HSA in most of

scenarios.

References

[1] Gopiya Naik S, Khatod DK, Sharma MP. Optimal allocationof combined DG and

capacitor for real power loss minimization in distribution networks. Int J Electr

Power Energy Syst 2013;53:967–73.

[2] Sathish Kumar K, Jayabarathi T. Power system reconfiguration and loss

minimization for an distribution systems using bacterial foraging

optimization algorithm. Int J Electr Power Energy Syst 2012;36(1):13–7 .

[3] Kalambe S, Agnihotri G. Loss minimization techniques used in distribution

network: bibliographical survey. Renew Sustain Energy Rev 2014;29:184–200.

[4] Nguyen TT, Truong AV. Distribution network reconfiguration for power loss

minimization and voltage profile improvement using cuckoo search algorithm.Int J Electr Power Energy Syst 2015;68:233–42.

[5] Merlin A, Back H. Search for a minimal loss operating spanning tree

configuration in an urban power distribution system. In: Proceeding 5th

power syst comput conf (PSCC), Cambridge, UK, vol. 1–18; 1975.

[6] Civanlar S, Grainger JJ, Yin H, Lee SSH. Distribution feeder reconfiguration for

loss reduction. IEEE Trans Power Deliv 1988;3(3):1217–23.

[7] Duan D-L, Ling X-D, Wu X-Y, Zhong B. Reconfiguration of distribution network

for loss reduction and reliability improvement based on an enhanced genetic

algorithm. Int J Electr Power Energy Syst 2015;64:88–95.

[8] Souza SSF, Romero R, Franco JF. Artificial immune networks Copt-aiNet and

Opt-aiNet applied to the reconfiguration problem of radial electrical

distribution systems. Electr Power Syst Res 2015;119:304–12.

[9] Esmaeilian HR, Fadaeinedjad R. Distribution system efficiency improvement

using network reconfiguration and capacitor allocation. Int J Electr Power

Energy Syst 2015;64:457–68.

[10] Hung DQ, Mithulananthan N, Bansal RC. An optimal investment planning

framework for multiple distributed generation units in industrial distribution

systems. Appl Energy 2014;124:62–72.

[11] Doagou-Mojarrad H, Gharehpetian GB, Rastegar H, Olamaei J. Optimal

placement and sizing of DG (distributed generation) units in distributionnetworks by novel hybrid evolutionary algorithm. Energy 2013;54:129–38.

[12] MurtyVVSN, Kumar A. Optimal placement of DG in radial distribution systems

based on new voltage stability index under load growth. Int J Electr Power

Energy Syst 2015;69:246–56.

[13] Mohamed IA, Kowsalya M. Optimal size and siting of multiple distributed

generators in distribution system using bacterial foraging optimization.

Swarm Evol Comput 2014;15:58–65.

[14] Ugranli F, Karatepe E. Multiple-distributed generation planning under load

uncertainty and different penetration levels. Int J Electr Power Energy Syst

2013;46(1):132–44.

[15] Kayal P, Chanda CK. Placement of wind and solar based DGs in distribution

system for power loss minimization and voltage stability improvement. Int J

Electr Power Energy Syst 2013;53:795–809.

[16] Ameli A, Shahab B, Farid K, Mahmood-Reza H. A multiobjective particle swarm

optimization for sizing and placement of DGs from DG Owner’s and

distribution company’s viewpoints. IEEE Trans Power Deliv 2014;29

(4):1831–40.

[17] Tan S, Xu JX, Panda SK. Optimization of distribution network incorporating

distributed generators: an integrated approach. IEEE Trans Power Syst

2013;28(3):2421–32.

[18] Rao RS, Ravindra K, Satish K, Narasimham SVL. Power loss minimization in

distribution system using network reconfiguration in the presence of

distributed generation. IEEE Trans Power Syst 2013;28(1):1–9.

[19] Mohamed Imran A, Kowsalya M, Kothari DP. A novel integration technique for

optimal network reconfiguration and distributed generation placement in

power distribution networks. Int J Electr Power Energy Syst 2014;63:461–72.

[20] Yang XS, Deb S. Cuckoo search via Lévy flights. In: 2009 world congress on

nature and biologically inspired computing, NABIC 2009 – proceedings; 2009.

p. 210–4.

[21] Balasubbareddy M, Sivanagaraju S, Suresh CV. Multi-objective optimization in

the presence of practical constraints using non-dominated sorting hybrid

cuckoo search algorithm. Eng Sci Technol Int J 2015;18(4):603–15.

[22] Abd Elazim SM, Ali ES. Optimal power system stabilizers design via cuckoo

search algorithm. Int J Electr Power Energy Syst 2016;75:99–107.

[23] Abdelaziz AY, Ali ES. Cuckoo search algorithm based load frequency controller

design for nonlinear interconnected power system. Int J Electr Power Energy

Syst 2015;73:632–43.[24] Dash P, Saikia LC, Sinha N. Comparison of performances of several FACTS

devices using Cuckoo search algorithm optimized 2DOF controllers in multi-

area AGC. Int J Electr Power Energy Syst 2015;65:316–24 .

[25] Chakravorty M, Das D. Voltage stability analysis of radial distribution

networks. Int J Electr Power Energy Syst 2001;23(2):129–35.

[26] Abdelaziz AY, Mohamed FM, Mekhamer SF, Badr MAL. Distribution system

reconfiguration using a modified Tabu search algorithm. Electr Power Syst Res

2010;80(8):943–53.

[27] Gupta N, Swarnkar A, Niazi KR, Bansal RC. Multi-objective reconfiguration of

distribution systems using adaptive genetic algorithm in fuzzy framework. IET

Gener Transm Distrib 2010;4(12):1288.

[28] Dolatdar E, Soleymani S, Mozafari B. A new distribution network

reconfiguration approach using a tree model. World Acad Sci Eng Technol

2009;3(10).

[29] Baran ME, Wu FF. Network reconfiguration in distribution systems for loss

reduction and load balancing. IEEE Trans Power Deliv 1989;4(2):1401–7.

[30] Chiang H-D, Jean-Jumeau R. Optimal network reconfigurations in distribution

systems: Part 2: solution algorithms and numerical results. IEEE Trans Power

Deliv 1990;5(3):1568–74.[31] Zhang D, Fu Z, Zhang L. An improved TS algorithm for loss-minimum

reconfiguration in large-scale distribution systems. Electr Power Syst Res

2007;77(5–6):685–94.

[32] Mohamed Imran A, Kowsalya M. A new power system reconfiguration scheme

for power loss minimization and voltage profile enhancement using Fireworks

Algorithm. Int J Electr Power Energy Syst 2014;62:312–22.

http://refhub.elsevier.com/S0142-0615(15)00582-7/h0005http://refhub.elsevier.com/S0142-0615(15)00582-7/h0005http://refhub.elsevier.com/S0142-0615(15)00582-7/h0005http://refhub.elsevier.com/S0142-0615(15)00582-7/h0010http://refhub.elsevier.com/S0142-0615(15)00582-7/h0010http://refhub.elsevier.com/S0142-0615(15)00582-7/h0010http://refhub.elsevier.com/S0142-0615(15)00582-7/h0010http://refhub.elsevier.com/S0142-0615(15)00582-7/h0015http://refhub.elsevier.com/S0142-0615(15)00582-7/h0015http://refhub.elsevier.com/S0142-0615(15)00582-7/h0015http://refhub.elsevier.com/S0142-0615(15)00582-7/h0020http://refhub.elsevier.com/S0142-0615(15)00582-7/h0020http://refhub.elsevier.com/S0142-0615(15)00582-7/h0020http://refhub.elsevier.com/S0142-0615(15)00582-7/h0030http://refhub.elsevier.com/S0142-0615(15)00582-7/h0030http://refhub.elsevier.com/S0142-0615(15)00582-7/h0035http://refhub.elsevier.com/S0142-0615(15)00582-7/h0035http://refhub.elsevier.com/S0142-0615(15)00582-7/h0035http://refhub.elsevier.com/S0142-0615(15)00582-7/h0040http://refhub.elsevier.com/S0142-0615(15)00582-7/h0040http://refhub.elsevier.com/S0142-0615(15)00582-7/h0040http://refhub.elsevier.com/S0142-0615(15)00582-7/h0040http://refhub.elsevier.com/S0142-0615(15)00582-7/h0045http://refhub.elsevier.com/S0142-0615(15)00582-7/h0045http://refhub.elsevier.com/S0142-0615(15)00582-7/h0045http://refhub.elsevier.com/S0142-0615(15)00582-7/h0050http://refhub.elsevier.com/S0142-0615(15)00582-7/h0050http://refhub.elsevier.com/S0142-0615(15)00582-7/h0050http://refhub.elsevier.com/S0142-0615(15)00582-7/h0055http://refhub.elsevier.com/S0142-0615(15)00582-7/h0055http://refhub.elsevier.com/S0142-0615(15)00582-7/h0055http://refhub.elsevier.com/S0142-0615(15)00582-7/h0055http://refhub.elsevier.com/S0142-0615(15)00582-7/h0060http://refhub.elsevier.com/S0142-0615(15)00582-7/h0060http://refhub.elsevier.com/S0142-0615(15)00582-7/h0060http://refhub.elsevier.com/S0142-0615(15)00582-7/h0065http://refhub.elsevier.com/S0142-0615(15)00582-7/h0065http://refhub.elsevier.com/S0142-0615(15)00582-7/h0065http://refhub.elsevier.com/S0142-0615(15)00582-7/h0065http://refhub.elsevier.com/S0142-0615(15)00582-7/h0070http://refhub.elsevier.com/S0142-0615(15)00582-7/h0070http://refhub.elsevier.com/S0142-0615(15)00582-7/h0070http://refhub.elsevier.com/S0142-0615(15)00582-7/h0070http://refhub.elsevier.com/S0142-0615(15)00582-7/h0075http://refhub.elsevier.com/S0142-0615(15)00582-7/h0075http://refhub.elsevier.com/S0142-0615(15)00582-7/h0075http://refhub.elsevier.com/S0142-0615(15)00582-7/h0080http://refhub.elsevier.com/S0142-0615(15)00582-7/h0080http://refhub.elsevier.com/S0142-0615(15)00582-7/h0080http://refhub.elsevier.com/S0142-0615(15)00582-7/h0080http://refhub.elsevier.com/S0142-0615(15)00582-7/h0085http://refhub.elsevier.com/S0142-0615(15)00582-7/h0085http://refhub.elsevier.com/S0142-0615(15)00582-7/h0085http://refhub.elsevier.com/S0142-0615(15)00582-7/h0090http://refhub.elsevier.com/S0142-0615(15)00582-7/h0090http://refhub.elsevier.com/S0142-0615(15)00582-7/h0090http://refhub.elsevier.com/S0142-0615(15)00582-7/h0090http://refhub.elsevier.com/S0142-0615(15)00582-7/h0095http://refhub.elsevier.com/S0142-0615(15)00582-7/h0095http://refhub.elsevier.com/S0142-0615(15)00582-7/h0095http://refhub.elsevier.com/S0142-0615(15)00582-7/h0105http://refhub.elsevier.com/S0142-0615(15)00582-7/h0105http://refhub.elsevier.com/S0142-0615(15)00582-7/h0105http://refhub.elsevier.com/S0142-0615(15)00582-7/h0105http://refhub.elsevier.com/S0142-0615(15)00582-7/h0110http://refhub.elsevier.com/S0142-0615(15)00582-7/h0110http://refhub.elsevier.com/S0142-0615(15)00582-7/h0115http://refhub.elsevier.com/S0142-0615(15)00582-7/h0115http://refhub.elsevier.com/S0142-0615(15)00582-7/h0115http://refhub.elsevier.com/S0142-0615(15)00582-7/h0115http://refhub.elsevier.com/S0142-0615(15)00582-7/h0120http://refhub.elsevier.com/S0142-0615(15)00582-7/h0120http://refhub.elsevier.com/S0142-0615(15)00582-7/h0120http://refhub.elsevier.com/S0142-0615(15)00582-7/h0125http://refhub.elsevier.com/S0142-0615(15)00582-7/h0125http://refhub.elsevier.com/S0142-0615(15)00582-7/h0130http://refhub.elsevier.com/S0142-0615(15)00582-7/h0130http://refhub.elsevier.com/S0142-0615(15)00582-7/h0130http://refhub.elsevier.com/S0142-0615(15)00582-7/h0130http://refhub.elsevier.com/S0142-0615(15)00582-7/h0135http://refhub.elsevier.com/S0142-0615(15)00582-7/h0135http://refhub.elsevier.com/S0142-0615(15)00582-7/h0135http://refhub.elsevier.com/S0142-0615(15)00582-7/h0135http://refhub.elsevier.com/S0142-0615(15)00582-7/h0140http://refhub.elsevier.com/S0142-0615(15)00582-7/h0140http://refhub.elsevier.com/S0142-0615(15)00582-7/h0140http://refhub.elsevier.com/S0142-0615(15)00582-7/h0145http://refhub.elsevier.com/S0142-0615(15)00582-7/h0145http://refhub.elsevier.com/S0142-0615(15)00582-7/h0150http://refhub.elsevier.com/S0142-0615(15)00582-7/h0150http://refhub.elsevier.com/S0142-0615(15)00582-7/h0150http://refhub.elsevier.com/S0142-0615(15)00582-7/h0155http://refhub.elsevier.com/S0142-0615(15)00582-7/h0155http://refhub.elsevier.com/S0142-0615(15)00582-7/h0155http://refhub.elsevier.com/S0142-0615(15)00582-7/h0160http://refhub.elsevier.com/S0142-0615(15)00582-7/h0160http://refhub.elsevier.com/S0142-0615(15)00582-7/h0160http://refhub.elsevier.com/S0142-0615(15)00582-7/h0160http://refhub.elsevier.com/S0142-0615(15)00582-7/h0160http://refhub.elsevier.com/S0142-0615(15)00582-7/h0160http://refhub.elsevier.com/S0142-0615(15)00582-7/h0155http://refhub.elsevier.com/S0142-0615(15)00582-7/h0155http://refhub.elsevier.com/S0142-0615(15)00582-7/h0155http://refhub.elsevier.com/S0142-0615(15)00582-7/h0150http://refhub.elsevier.com/S0142-0615(15)00582-7/h0150http://refhub.elsevier.com/S0142-0615(15)00582-7/h0150http://refhub.elsevier.com/S0142-0615(15)00582-7/h0145http://refhub.elsevier.com/S0142-0615(15)00582-7/h0145http://refhub.elsevier.com/S0142-0615(15)00582-7/h0140http://refhub.elsevier.com/S0142-0615(15)00582-7/h0140http://refhub.elsevier.com/S0142-0615(15)00582-7/h0140http://refhub.elsevier.com/S0142-0615(15)00582-7/h0135http://refhub.elsevier.com/S0142-0615(15)00582-7/h0135http://refhub.elsevier.com/S0142-0615(15)00582-7/h0135http://refhub.elsevier.com/S0142-0615(15)00582-7/h0130http://refhub.elsevier.com/S0142-0615(15)00582-7/h0130http://refhub.elsevier.com/S0142-0615(15)00582-7/h0130http://refhub.elsevier.com/S0142-0615(15)00582-7/h0125http://refhub.elsevier.com/S0142-0615(15)00582-7/h0125http://refhub.elsevier.com/S0142-0615(15)00582-7/h0120http://refhub.elsevier.com/S0142-0615(15)00582-7/h0120http://refhub.elsevier.com/S0142-0615(15)00582-7/h0120http://refhub.elsevier.com/S0142-0615(15)00582-7/h0115http://refhub.elsevier.com/S0142-0615(15)00582-7/h0115http://refhub.elsevier.com/S0142-0615(15)00582-7/h0115http://refhub.elsevier.com/S0142-0615(15)00582-7/h0110http://refhub.elsevier.com/S0142-0615(15)00582-7/h0110http://refhub.elsevier.com/S0142-0615(15)00582-7/h0105http://refhub.elsevier.com/S0142-0615(15)00582-7/h0105http://refhub.elsevier.com/S0142-0615(15)00582-7/h0105http://refhub.elsevier.com/S0142-0615(15)00582-7/h0095http://refhub.elsevier.com/S0142-0615(15)00582-7/h0095http://refhub.elsevier.com/S0142-0615(15)00582-7/h0095http://refhub.elsevier.com/S0142-0615(15)00582-7/h0090http://refhub.elsevier.com/S0142-0615(15)00582-7/h0090http://refhub.elsevier.com/S0142-0615(15)00582-7/h0090http://refhub.elsevier.com/S0142-0615(15)00582-7/h0085http://refhub.elsevier.com/S0142-0615(15)00582-7/h0085http://refhub.elsevier.com/S0142-0615(15)00582-7/h0085http://refhub.elsevier.com/S0142-0615(15)00582-7/h0080http://refhub.elsevier.com/S0142-0615(15)00582-7/h0080http://refhub.elsevier.com/S0142-0615(15)00582-7/h0080http://refhub.elsevier.com/S0142-0615(15)00582-7/h0080http://refhub.elsevier.com/S0142-0615(15)00582-7/h0075http://refhub.elsevier.com/S0142-0615(15)00582-7/h0075http://refhub.elsevier.com/S0142-0615(15)00582-7/h0075http://refhub.elsevier.com/S0142-0615(15)00582-7/h0070http://refhub.elsevier.com/S0142-0615(15)00582-7/h0070http://refhub.elsevier.com/S0142-0615(15)00582-7/h0070http://refhub.elsevier.com/S0142-0615(15)00582-7/h0065http://refhub.elsevier.com/S0142-0615(15)00582-7/h0065http://refhub.elsevier.com/S0142-0615(15)00582-7/h0065http://refhub.elsevier.com/S0142-0615(15)00582-7/h0060http://refhub.elsevier.com/S0142-0615(15)00582-7/h0060http://refhub.elsevier.com/S0142-0615(15)00582-7/h0060http://refhub.elsevier.com/S0142-0615(15)00582-7/h0055http://refhub.elsevier.com/S0142-0615(15)00582-7/h0055http://refhub.elsevier.com/S0142-0615(15)00582-7/h0055http://refhub.elsevier.com/S0142-0615(15)00582-7/h0050http://refhub.elsevier.com/S0142-0615(15)00582-7/h0050http://refhub.elsevier.com/S0142-0615(15)00582-7/h0050http://refhub.elsevier.com/S0142-0615(15)00582-7/h0045http://refhub.elsevier.com/S0142-0615(15)00582-7/h0045http://refhub.elsevier.com/S0142-0615(15)00582-7/h0045http://refhub.elsevier.com/S0142-0615(15)00582-7/h0040http://refhub.elsevier.com/S0142-0615(15)00582-7/h0040http://refhub.elsevier.com/S0142-0615(15)00582-7/h0040http://refhub.elsevier.com/S0142-0615(15)00582-7/h0035http://refhub.elsevier.com/S0142-0615(15)00582-7/h0035http://refhub.elsevier.com/S0142-0615(15)00582-7/h0035http://refhub.elsevier.com/S0142-0615(15)00582-7/h0030http://refhub.elsevier.com/S0142-0615(15)00582-7/h0030http://refhub.elsevier.com/S0142-0615(15)00582-7/h0020http://refhub.elsevier.com/S0142-0615(15)00582-7/h0020http://refhub.elsevier.com/S0142-0615(15)00582-7/h0020http://refhub.elsevier.com/S0142-0615(15)00582-7/h0015http://refhub.elsevier.com/S0142-0615(15)00582-7/h0015http://refhub.elsevier.com/S0142-0615(15)00582-7/h0010http://refhub.elsevier.com/S0142-0615(15)00582-7/h0010http://refhub.elsevier.com/S0142-0615(15)00582-7/h0010http://refhub.elsevier.com/S0142-0615(15)00582-7/h0005http://refhub.elsevier.com/S0142-0615(15)00582-7/h0005http://refhub.elsevier.com/S0142-0615(15)00582-7/h0005http://-/?-http://-/?-

Documents

1-s2.0-S0142061515005827-main.pdf