Upload
phung-tuan
View
214
Download
0
Embed Size (px)
Citation preview
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.
804 T.T. Nguyen et al. / Electrical Power and Energy Systems 78 (2016) 801–815
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.
T.T. Nguyen et al./ Electrical Power and Energy Systems 78 (2016) 801–815 805
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
If hosts nest Xi is radial configuration then
Evaluate fitness function to choose new Xbest d
Else
Fitness function of Xi = inf
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
If hosts nest Xi is radial configuration then
Evaluate fitness function to choose new Xbest d and Gbest
Else
Fitness function of Xi = inf
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.
806 T.T. Nguyen et al. / Electrical Power and Energy Systems 78 (2016) 801–815
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
T.T. Nguyen et al./ Electrical Power and Energy Systems 78 (2016) 801–815 807
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
Minimum VSI 0.7878 – –
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
Minimum voltage (p.u.) 0.9778 0.9680 0.9670
Minimum VSI 0.9118 – –
DG installation after reconfiguration
(Scenario 4)
Switches opened 7, 14, 9, 32, 28 7, 14, 9, 32, 28 7, 14, 9, 32, 37
Size of DG in MW (Bus number) 1.7536 (29) 0.5996 (32) 0.2686 (32)
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
Minimum voltage (p.u.) 0.9802 0.9612 0.9479
Minimum VSI 0.9264 – –
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 –
Minimum VSI 0.9354 – –
Simultaneous Reconfiguration and DG
installation (Scenario 6)
Switches opened 7, 10, 13, 32, 27 7, 14, 11, 32, 28 7, 14, 10, 32, 28
Size of DG in MW (Bus number) 0.4263 (32) 0.5367 (32) 0.5258 (32)
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
Minimum voltage (p.u.) 0.9786 0.9713 0.9700
Minimum VSI 0.9202 – –
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)
Power loss (kW) 53.21 – –
% Loss reduction 73.75 – –
Minimum voltage (p.u.) 0.9806 – –
Minimum VSI 0.9318 – –
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.
808 T.T. Nguyen et al. / Electrical Power and Energy Systems 78 (2016) 801–815
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.
T.T. Nguyen et al./ Electrical Power and Energy Systems 78 (2016) 801–815 809
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
Performance analysis of proposed method on the 69-bus system.
Scenario Item Proposed ACSA FWA [19] HSA [18]Base case (Scenario 1) Switches opened 69, 70, 71, 72, 73 – –
Power loss (kW) 224.89 – –
Minimum voltage (p.u.) 0.9092 – –
Minimum VSI 0.6859 – –
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
Minimum voltage (p.u.) 0.9495 0.9495 0.9428
Minimum VSI 0.8414 – –
Only DG installation (Scenario 3) Switches opened 69, 70, 71, 72, 73 69, 70, 71, 72, 73
Size of DG in MW (Bus number) 0.6022 (11) 0.4085 (65) 0.1018 (65)
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
Size of DG in MW (Bus number) 1.7254 (61) 1.0014 (61) 1.0666 (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
Minimum voltage (p.u.) 0.9870 0.9720 0.9619
Minimum VSI 0.9390 – –
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 –
Minimum VSI 0.9260 – –
Simultaneous Reconfiguration
and DG installation (Scenario
6)
Switches opened 69, 70, 12, 58, 61 69, 70, 13, 55, 63 69, 17, 13, 58, 61
Size of DG in MW (Bus number) 1.7496 (61) 1.1272 (61) 1.0666 (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
Minimum voltage (p.u.) 0.9873 0.9796 0.9736
Minimum VSI 0.9403 – –
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)
Power loss (kW) 37.02 – –
% Loss reduction 83.54 – –
Minimum voltage (p.u.) 0.9869 – –
Minimum VSI 0.9433 – –
810 T.T. Nguyen et al. / Electrical Power and Energy Systems 78 (2016) 801–815
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
Fig. 12. Comparison of node voltages of 69-bus system.
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
Fig. 13. Comparison of VSI-nodes of 69-bus system.
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.
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. 14. Comparison of 69-bus system indices for scenarios in fitness function.
T.T. Nguyen et al./ Electrical Power and Energy Systems 78 (2016) 801–815 811
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
Performance analysis of proposed method on the 119-bus system.
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
Power loss (kW) 855.04
% Loss reduction 32.86
Minimum voltage (p.u.) 0.9298
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)
Power loss (kW) 648.10
% Loss reduction 49.11
Minimum voltage (p.u.) 0.9515
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
Size of DG in MW (Bus
number)
1.7145 (111)
1.7565 (96)5 (65)
Power loss (kW) 631.19
% Loss reduction 50.43
Minimum voltage (p.u.) 0.9526
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
Size of DG in MW (Bus
number)
3.2664 (71)
3.1203 (109)
2.86267 (50)
Power loss (kW) 613.79
% Loss reduction 51.80
Minimum voltage (p.u.) 0.9608
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
Size of DG in MW (Bus
number)
2.9585 (75)
0.1924 (76)
1.3397 (77)
Power loss (kW) 682.09
% Loss reduction 46.44
Minimum voltage (p.u.) 0.9298
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
Size of DG in MW (Bus
number)
2.5331 (50)
3.6819 (109)
3.7043 (73)
Power loss (kW) 586.24
% Loss reduction 53.96
Minimum voltage (p.u.) 0.9644
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
T.T. Nguyen et al./ Electrical Power and Energy Systems 78 (2016) 801–815 813
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
Fig. 16. Comparison of node voltages of 119-bus system.
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
Fig. 17. Comparison of VSI-nodes of 119-bus system.
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.
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. 18. Comparison of performance of ACSA in six scenarios for minimization of the 119-bus system.
814 T.T. Nguyen et al. / Electrical Power and Energy Systems 78 (2016) 801–815
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.
T.T. Nguyen et al./ Electrical Power and Energy Systems 78 (2016) 801–815 815
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://-/?-