Upload
others
View
0
Download
0
Embed Size (px)
Citation preview
A
Igm
KaQ2bQ3Q4
a
ARRAA
KRP(DSH
1
Q5ctPwbtapcmtcd
h1
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
ARTICLE IN PRESSG ModelSOC 3714 1–22
Applied Soft Computing xxx (2016) xxx–xxx
Contents lists available at ScienceDirect
Applied Soft Computing
j ourna l h o mepage: www.elsev ier .com/ locate /asoc
ntegrated approach of network reconfiguration with distributedeneration and shunt capacitors placement for power lossinimization in radial distribution networks
. Muthukumar (Assistant Professor) a,∗, S. Jayalalitha (Associate Dean) b
EEE/SEEE, SASTRA University, Tirumalaisamudram, Thanjavur, Tamilnadu, IndiaEIE/SEEE, SASTRA University, Tirumalaisamudram, Thanjavur, Tamilnadu, India
r t i c l e i n f o
rticle history:eceived 5 January 2016eceived in revised form 15 July 2016ccepted 17 July 2016vailable online xxx
eywords:adial distribution network (RDN)article artificial bee colony algorithmPABC)
a b s t r a c t
This article presents the significance of efficient hybrid heuristic search algorithm(HS-PABC) based onHarmony search algorithm (HSA) and particle artificial bee colony algorithm (PABC) in the context ofdistribution network reconfiguration along with optimal allocation of distributed generators and shuntcapacitors. The premature and slow convergence over multi model fitness landscape is the main limitationin standard HSA. In the proposed hybrid algorithm the harmony memory vector of HSA are intelligentlyenhanced through PABC algorithm during the optimization process to reach the optimal solution withinthe search space. In hybrid approach, the exploration ability of HSA and the exploitation ability of PABCalgorithm are integrated to blend the potency of both algorithms. The box plot and Wilcoxon rank sumtest are used to show the quality of the solution obtained by hybrid HS-PABC with respect to HSA.The
istributed generator (DG)hunt capacitorsarmony search algorithm (HSA)
computational results prove the integrated approach of the network reconfiguration problem along withoptimal placement and sizing of DG units and shunt capacitors as an efficient approach towards theobjective. The results obtained on 69 and 118 node network by proposed method and the standard HSAreveals the powerfulness of the proposed approach which guarantees to achieve global optimal solutionwith less iteration.
© 2016 Elsevier B.V. All rights reserved.
39
40
41
42
43
44
45
46
47
48
49
50
51
52
. Introduction
The locality of electrical energy generation is far away from theonsumer loads connected with lengthy feeder lines leads to addi-ional power loss in the transmission and distribution network.ower loss reduction is achieved by reconfiguring the existing net-ork topology and by installing fixed/switched shunt capacitor
anks and distributed generation units (DG) in close proximity tohe consumer loads in transmission and distribution networks. Thellocation of such sources has numerous advantages such as post-onement for investing new transmission and distribution networkonstruction, reduction in power loss, bus voltage profile enhance-ent. Prior to the implementation of loss reduction techniques in
he distribution network; there is a necessary to investigate their
Please cite this article in press as: M. K., J. S., Integrated approand shunt capacitors placement for power loss minimization inhttp://dx.doi.org/10.1016/j.asoc.2016.07.031
onsequence, such as power loss, bus voltage magnitude, harmonicistortion and system voltage stability. A suitable planning method
∗ Corresponding author.E-mail address: [email protected] (M. K.).
ttp://dx.doi.org/10.1016/j.asoc.2016.07.031568-4946/© 2016 Elsevier B.V. All rights reserved.
53
54
55
56
must be implemented to get the benefits of integrating the DG unitsand shunt capacitors into the distribution networks.
The network reconfiguration of the RDN is the method of chang-ing the topological structure of the network by opening and closingof sectionalizing and tie switches to achieve optimal topology withminimum power loss. During the reconfiguration process, the sys-tem radiality should be maintained with all loads connected to thenetwork. In recent years, a noticeable research work has been car-ried out for loss reduction using network reconfiguration problem.Since there are numerous candidate switching combinations to findthe optimal topology of the network, the reconfiguration prob-lem is modeled as combinatorial, non-differentiable, constrainedoptimization problem. The discrete nature of sectionalizing and tieswitches along with radiality constraints avert the application ofclassical optimization methodologies. So there has been a growinginterest in various population based heuristic search algorithmssuch as Artificial Immune (AIS) Systems [1], Modified particleswarm optimization [2], Binary group search optimization [3],
ach of network reconfiguration with distributed generation radial distribution networks, Appl. Soft Comput. J. (2016),
Adapted ant colony [4], Fireworks algorithm [5], Harmony search[6] algorithm. A fuzzy multi objective network reconfigurationmethodology for radial distribution systems has been proposed in
57
58
59
ARTICLE ING ModelASOC 3714 1–22
2 M. K., J. S. / Applied Soft Compu
Nomenclature
QDGi Reactive power injection of ith DG unit
p.f DGmax Maximum allowable power factor of DG unit
p.f DGmin Minimum allowable power factor of DG unit
p.f DGi Power factor of ith DG unit
nbus Total number of nodes in the RDNnb Total branches in the RDNnc Total number of capacitors to be installed in RDNIi,i + 1 Current flow between ith branch and i + 1th branchIi,i + 1max Allowable maximum permissible current at branch
i + 1IBmax(i) Maximum allowable branch current
IWcompBmax(i) Branch current flow in ith branch with compensa-
tionPw/comp
loss(i,i+1) Real ower loss in the branch between nodes i andi + 1 with DG units, shunt capacitor
Pw/compTloss
Total power loss in the RDN with DG units, shuntcapacitor
|Vi| Bus voltage magnitude of ith busVspec
min Specified lower bound of bus voltage of the RDNVspec
max Specified upper bound of bus voltage of RDNPDG
max Maximum size of DG unit in kilowattsPDG
min Minimum size of DG unit in kilowattsQcL Sum of total reactive demand of the RDNQcj Reactive power injection by the jth shunt capacitor
[uaprfcautmimotfiut[op
csmslptsttohA
60
61
62
63
64
65
66
67
68
69
70
71
72
73
74
75
76
77
78
79
80
81
82
83
84
85
86
87
88
89
90
91
92
93
94
95
96
97
98
99
100
101
102
103
104
105
106
107
108
109
110
111
112
113
114
115
116
117
118
119
120
121
122
123
124
125
126
127
128
129
130
131
132
133
134
135
136
137
138
139
140
141
142
143
144
145
146
147
148
149
150
151
152
153
154
155
156
PiDG Size of ith DG unit in kilowatts
7]. Tabu search approach has been proposed for network reconfig-ration in [8–10]. Mathematical illustration of radiality constraintsssociated with the radial distribution network reconfigurationroblem has been addressed in [11]. Optimal network reconfigu-ation along with the placement of distribution generation unitsor power loss minimization has been proposed in [12,13]. Asompared with classical optimization techniques, meta-heuristiclgorithms are capable to achieve a near optimal solution for DGnits and shunt capacitors sizing and placement problem in dis-ribution networks. Artificial Bee Colony (ABC) algorithm which
imics the foraging nature of honey bees swarm has been proposedn [14] to find the best possible sizing of static capacitors. Various
odels and methods applied for the optimal DG allocation, impactf dispatchable and non-dispatchable type DG units integration inhe modern distribution system, current and future trends in thiseld has been addressed in [15,16]. Analytical approach based DGnits sizing and its optimal operating power factor for differentypes of DG unit’s for power loss reduction has been addressed in17]. An analytical approach for simultaneous allocation and sizingf DG units and shunt capacitor to achieve power loss reduction isroposed in [18].
In [19] a heuristic algorithm, mimicking the improvisation pro-ess of musical instruments to get a pleasant harmonious melodyo called harmony search algorithm (HSA) is proposed. The perfor-ance of the developed algorithm is demonstrated with a traveling
alesman problem and minimum-cost pipe network design prob-em. In [20] an improved harmony search algorithm (IHSA) isroposed which intelligently generates new solution vector forhe best solution and convergence of HSA. HSA is a stochasticearch procedure which does not need any derivative informationo solve the complex combinatorial optimization problem. In [21]
Please cite this article in press as: M. K., J. S., Integrated approand shunt capacitors placement for power loss minimization inhttp://dx.doi.org/10.1016/j.asoc.2016.07.031
he discrete search strategy of HSA is utilized for structural sizeptimization problem with discrete design variables. In [22,23] aybrid heuristic search approach is proposed which makes use ofBC algorithm and its variants to enhance the solution vector in the
PRESSting xxx (2016) xxx–xxx
HS algorithm and the simulation results are compared with the har-mony search algorithm (HSA), improved harmony search algorithm(IHSA), global harmony search algorithm (GHSA) and self adaptiveglobal harmony search (SGHS) algorithm. The parameters of hybridalgorithm and its impact has been studied with uniform designexperiments optimization problems. In [24,25] the optimal designof water distribution networks is addressed using HSA incorpo-rated particle swarm algorithm (PSO) and the result outcomes arebetter than the genetic algorithm, simulated annealing and Tabusearch technique. In [26] a self-adaptive global best harmony searchalgorithm (SGHS) is proposed which utilize a new harmony mem-ory enhancement strategy by dynamic adaptation of HMCR andPAR, distance bandwidth (BW) as learning mechanism to balancethe exploration and exploitation ability. In [27] a hybrid heuris-tic algorithm is proposed which makes use of sequential quadraticprogramming technique (HSA-SQL) to accelerate the local searchability and to get better accuracy in HSA solutions. To exhibitthe effectiveness and sturdiness of the proposed hybrid HSA-SQLalgorithm, various benchmark engineering optimization problemsare taken into consideration. The intelligent honey bee foragingbehavior of bee swarm is utilized in [28–30] to obtain the optimalsolution in multi-dimensional numerical optimization problems.The simulation results outperform the results obtained with theother meta-heuristic algorithms such as a particle swarm opti-mization algorithm, differential evolution algorithm and geneticalgorithms. In calculus based methods, the optimal solution isobtained by using derivatives which is suitable only for continuous-valued functions rather than discrete-valued functions. In [31,32]a new HS algorithm is proposed for solving engineering optimiza-tion problems with continuous as well as discrete design variables.In [33,34] the overview of the recent applications of HSA, whichuses a ‘probabilistic-gradient’ to select the neighboring values ofdecision variables is addressed. It is an efficient meta-heuristicoptimization tool for practitioners to solve complex optimizationparadigms such as construction, telecommunications, engineering,health and energy, and robotics. The HSA provides probabilistic-gradient based search to get the local or global optimal solutioninstead of mathematical gradient as in conventional optimizationtechniques. In [35–38] a HSA based optimal solution is obtained forcomplex optimization problems like scheduling of multiple damsystems, broadcast scheduling in packet radio networks, estima-tion of the success of companies and vehicle routing. In [39] a hybridgrouping HSA is proposed for the multiple-type access node loca-tion problem to determine the optimum location. In [40] a hybridapproach is used to deploy 24-h medical emergency resources bycombining the HSA with the grouping encoding concept to repairinfeasible solutions. In [41] a multi-objective harmony search forurban road network reconfiguration problem to offer near-optimalsolution to improve the vehicles mobility is proposed. In [42] aquasi-oppositional harmony search algorithm is proposed to inves-tigate the optimal controller gains to enhance the performance ofAutomatic Generation Control (AGC) of the power system. A newhybrid PSO algorithm has been addressed in [43] to augment theexploration and exploitation capability by introducing the globaldimension selection strategy using HSA and validated with the PSOvariants, and other meta-heuristic algorithms. In [44] a new self-adaptive HS-PSO search algorithm is proposed with an effectiveinitialization scheme by utilizing the PSO algorithm to improve thesolution quality of the initial harmony memory in the HSA. A newself-adaptive adjusting method for control parameters PAR and BWis designed to accelerate the convergence rate and solution accu-racy of the proposed algorithm. The poor exploitation ability of the
ach of network reconfiguration with distributed generation radial distribution networks, Appl. Soft Comput. J. (2016),
ABC algorithm makes an issue of slow convergence in solving non-linear and constrained optimization nature of engineering designproblems. To overcome these insufficiencies, a modified version ofthe ABC algorithm is suggested in [45–49] by incorporating adap-
157
158
159
160
ING ModelA
ompu
ta
acccoathlassct
btwmHHgtilwbcpa
pcHlo
2
iloiQ6p
M
w
m
2
tpts
R
Nb = Nbus − 1
where Nb is the number of branches, Nbus is the number of nodesin the RDN.
161
162
163
164
165
166
167
168
169
170
171
172
173
174
175
176
177
178
179
180
181
182
183
184
185
186
187
188
189
190
191
192
193
194
195
196
197
198
199
200
201
202
203
204
205
206
207
208
209
210
211
212
213
214
215
216
217
218
219
220
221
222
223
224
225
226
227
228
229
230
231
232
233
234
235
236
237
238
239
240
241
242
243
244
245
246
247
248
249
250
251
252
ARTICLESOC 3714 1–22
M. K., J. S. / Applied Soft C
ive inertia weights and acceleration coefficient parameters of PSOlgorithms.
The combined approach of network reconfiguration problemlong with the placement of distributed generators and shuntapacitors is complex combinatorial problems which involves dis-rete (status of sectionalizing and tie switches) and continuousontrol variables (size of DG units and shunt capacitors). To find theptimal topology all the feasible radial topologies of the networkre to be evaluated. Tracing such optimal topology is computa-ionally expensive and NP- hard in nature. Solving such type ofard optimization problems using meta-heuristic techniques will
ead to premature convergence and local optima. Improper bal-nce between intensification and diversification to find the optimalolution is the major issue in meta-heuristic techniques. This neces-itates the hybridization of meta-heuristics which makes use ofomplementary characteristic of different optimization strategieso get the improved solution.
In this article, hybridization of HSA with PSO embedded artificialee colony algorithm (HS-PABC) has been proposed and utilizedo provide the optimal network reconfiguration problem alongith the placement of DG units and shunt capacitors to realizeinimization of real power loss at distinct load levels. In hybrid,S-PABC approach, the global searching ability (exploration) ofSA and the local searching (exploitation) ability of PABC are inte-rated to harmoniously enhance the HS algorithm performance. Inhis way, the strength of local searching ability of standard HSA ismproved to find out the best solution quickly and ride over theocal optima. In this work, different test scenarios are considered
ith an aim to enumerate the benefits of RDN through graph theoryased network reconfiguration along with the allocation of shuntapacitors and DG units. Simulation results point out that the pro-osed hybrid HS-PABC algorithm is proficient than standard HSAlgorithm.
This article is structured as: In Section 2 the formulation of theroblem and graph theory based network reconfiguration is dis-ussed. Outline of standard HS algorithm and the proposed hybridS-PABC algorithm is summarized in Section 3. In Section 4, simu-
ation results and discussion are presented. In Section 5 conclusionsf the work are presented.
. Problem formulation
A multi objective optimization model has been proposed whichncludes technical factors such as real power loss reduction, lineoading reduction and voltage profile improvement. The multibjective proposition is considered simultaneously using weight-
ng factors for obtaining maximum benefits as in Eq. (1). All theerformance indices are given equal weightage.
in.fitness = W1(RPLRI) + W2(MBCCLI) + W3(VDI) (1)
here∑3
i=1wi = 1.0 �wi� [0,1] .A set of performance improvement indices is described mathe-
atically as stated below to enumerate the technical benefits.
.1. Real power loss reduction index (RPLRI)
The objective of real power loss reduction index is to quan-ify the power loss reduction in the distribution network. The realower loss reduction index (RPLRI) is defined as the ratio betweenhe real power loss of the network with and without DG units and
Please cite this article in press as: M. K., J. S., Integrated approand shunt capacitors placement for power loss minimization inhttp://dx.doi.org/10.1016/j.asoc.2016.07.031
hunt capacitor compensation as in Eq. (2),
PLRI = (plosswcomp
plosswocomp ) (2)
PRESSting xxx (2016) xxx–xxx 3
2.2. Maximum branch current capacity limit index (MBCCLI)
This index is used to reflect the improvement in reserve capacitylimit of the feeder lines. It is the ratio of current flow through thefeeder lines after compensation and the maximum current carryinglimit of that feeder branch as in Eq. (3),
MBCCLI = max(IBwcomp
(i)
IBmax(i)) (3)
2.3. Voltage deviation index (VDI)
The optimal network reconfiguration along with DG units andshunt capacitor placement provides the reduction in bus voltagedeviation. The bus voltage deviation index (VDI) can be defined asin Eq. (4)
VDI = max(V1 − Vi
V1)∀i = 1, 2, .....nbus whereV1
is taken as 1.0p.u (4)
2.4. Constraints associated with the objective function
The nominal bus voltage limit, thermal limit of the feeder lines,maximum real and reactive power injection by the DG units andshunt capacitors, radial structure of the distribution network areconsidered as constraints in Eqs. (5)–(10).
|Vspecmin
| ≤ |Vi| ≤ |Vspecmax | (5)
0.9 ≤ |Vi| ≤ 1.1 i = 1, 2, ....Nbus
|Ii,i+1| ≤ |Ii,i+1max|, i = 1, 2, ....Nb
(6)
∑nc
j=1Qcj ≤ 1.0
∑nl
j=1QcL (7)
The DG units are modeled as controllable synchronous machinebased biomass plants with active and reactive power injectioncapabilities at an operating power factor of 0.85 lagging [17]. Thereal power output (kW) range of DG units sizing is selected as10–50% of the total active power demand of the RDN.
PDGmin ≤ PDG
i ≤ PDGmax (8)
where PDGmin = 0.1
n∑i=2
PDGi and PDG
Max = 0.5n∑
i=2
PDGi
QDGi = PDG
i tan(cos−1(p.f)) (9)
p.fDGmax ≥ p.fDG
i ≥ p.fDGmin (10)
Radiality constraints:
ach of network reconfiguration with distributed generation radial distribution networks, Appl. Soft Comput. J. (2016),
2.5. Spanning tree approach for network reconfiguration
A directed graph representation of the radial distribution net-work is utilized to form the adjacency matrix (AM) with ‘1’s and
253
254
ING ModelA
4 ompu
‘nodwig
sntormstw(erp
iiEif
A
Fmootiao
A
t
wb
(
ocfiSw
255
256
257
258
259
260
261
262
263
264
265
266
267
268
269
270
271
272
273
274
275
276
277
278
279
280
281
282
283
284
285
286
287
288
289
290
291
292
293
294
295
296
297
298
299
300
301
302
303
304
305
306
307
308
309
310
311
312
313
314
315
316
317
318
319
320
321
322
323
324
325
326
327
328
329
330
331
332
333
334
335
336
337
338
339
340
ARTICLESOC 3714 1–22
M. K., J. S. / Applied Soft C
0’s representing the presence and absence of direct edge betweenodes. The node isolation is identified by determining the columnsf the adjacency matrix with only zeros except the root node and inegree of each node is limited to one by determining the columnsith only one nonzero element. The formation of the loop dur-
ng reconfiguration process is identified using MATLAB functionraphisdag (AM).
In this proposed work, the graph data structure is used to con-truct the RDN as spanning tree. The vertices representing theodes of RDN connected with edges (branches) are used to formhe directed graph. The load flow algorithm developed is basedn the formation of BIBC and BCBV matrices [50]. The BIBC matrixelates the bus current injection with branch current where as BCBV
atrix relates the branch current with bus voltage which repre-ents the line impedance of the RDN. In each switching operation,he adjacency matrix and BCBV matrix are changed in accordanceith the change in the system structure. The depth-first search
DFS) discovery order is utilized for traversing the graph fromach node. Based on the discovery order from each node, the cor-esponding BIBC matrix is generated by assigning ‘1′ to discoveraths.
The hypothetical 6 node RDN with sectionalizing and tie switchs shown in Fig. 1(a). The adjacency matrix representation is shownn Eq. (18). Let E be a set of direct edges {(1,2),(2,3),(3,4),(4,5),(3,6)}.ach element in the adjacency matrix AMij = 1 if (i,j)C-- E and AMij = 0
f (i,j) /∈ E. The adjacency matrix before closing the tie switch isormulated as in Eq. (11).
M = [
0
00000
1
00000
0
10000
0
01000
0
00100
0
01000
] (11)
ig. 1(b) represents the directed graph generated using adjacencyatrix represented in (11). The RDN structure modification is
btained by closing the tie switch between the node 5 & 6 andpening the sectionalizing switch between the node 4 and 5 andhe corresponding modified adjacency matrix obtained is shownn Eq. (12). Fig. 1(c) represents the directed graph generated usingdjacency matrix represented in Eq. (12) for the modified structuref the RDN.
M = [
0
00000
1
00000
0
10000
0
01000
0
00001
0
01000
] (12)
The steps involved in solving the optimal network reconfigura-ion problem of RDN are lined as follows:
Step 1: Generate the adjacency matrix (AM) of the given RDNith ‘1’ and ‘0’ representing the presence and absence of direct edge
etween the nodes.
Please cite this article in press as: M. K., J. S., Integrated approand shunt capacitors placement for power loss minimization inhttp://dx.doi.org/10.1016/j.asoc.2016.07.031
Step2: Generate the directed graph using function, bio-graphbg) from the adjacency matrix (AM).
Step 3: Traverse the directed graph in depth first search (DFS)rder using function traverse (bg, S) where S = 1,2. . ..n to obtain dis-overy vector. The power flow path matrix [i.e., BIBC] is formed bylling ‘1’ in position (i,j) where ith row corresponds to each value of
and jth column corresponds to each element in discovery vectorith respect to S.
PRESSting xxx (2016) xxx–xxx
Step 4: Obtain the branch current [BC] by multiplying BIBCmatrix obtained from step 3 with bus current injection [I]
[BC] = [BIBC ∗ I] (13)
Step 5: The distribution load flow matrix (DLF) is obtained bymultiplying BCBV and BIBC matrices which relates the bus currentinjection and bus voltages and it is expressed as,
�V = [BCBV] ∗ [BIBC] ∗ [I] = [DLF ∗ I] (14)
The bus current injection in i + 1th node at kth iteration can beobtained as follows,
Iki+1 =
√P2
i+1 + Q2i+1
|Vi|2(15)
The change in bus voltage at i + 1th node at k + 1th iteration can beobtained as follows,
�VK+1i+1 = [DLF ∗ IKi+1] (16)
VK+1I+1 = V0 + �VK+1
I+1 (17)
In this similar fashion, the bus voltages are updated during iter-ative steps. The iterative steps are terminated once the requiredtolerance limit is reached.
3. Evolutionary algorithms
3.1. Overview of harmony search algorithm
Harmony search algorithm (HSA) was proposed by Geem et al.[19]. The HSA algorithm requires fewer mathematical calcula-tions and it does not require any derivative information due to itsstochastic searching nature. Various advantages of HS algorithmare (a) it does not require differential gradients, thus it can be suit-able for discontinuous as well as continuous functions; (b) it canhave the capability to handle discrete as well as continuous deci-sion variables [32]; (c) it does not require initial value setting forthe decision variables.
HMCR, PAR and BW are considered as tuning parameters inHSA. The computational procedures of HSA are a) random initial-ization of decision variables in the harmony memory vector (HMV)b) Improvisation of HMV values by updating the solutions by mem-ory consideration c) pitch adjustment and random re-initializationprocess. Repeat the improvisation process in the HMV until thetermination criteria.
Updating of harmony memory vector is described below.
ach of network reconfiguration with distributed generation radial distribution networks, Appl. Soft Comput. J. (2016),
(18)HM is pitch adjusted or modified slightly with a probability of
PAR as,
341
342
343
ARTICLE IN PRESSG ModelASOC 3714 1–22
M. K., J. S. / Applied Soft Computing xxx (2016) xxx–xxx 5
hypoth
(w
3a
tmttbtttprt
ts
344
345
346
347
348
349
350
351
352
353
354
355
356
357
358
359
360
361
362
363
364
365
366
367
368
369
370
371
372
373
374
375
376
Fig. 1. (a) Example hypothetical 6 node RDN with directed graph of the
19)where “BW” represents a random distance bandwidth. In thisay the HMV is improved iteratively.
.2. PSO embedded artificial bee colony algorithm (PABClgorithm)
Particle swarm optimization (PSO) is one of the recent stochas-ic algorithm based on the idea of swarm intelligence of social
odels such as bird flocking, fish schooling [50]. In PSO, simul-aneous co-existence of multiple candidate solutions flies withinhe problem search space. Each particle adjusts its flying positionased on its own experience as well as the best flying position ofhe neighboring particles to reach the optimal solution point. Inhis way each particle retains its best position (Pbest—local searchhrough its own experience) and its neighboring particle’s best
Please cite this article in press as: M. K., J. S., Integrated approand shunt capacitors placement for power loss minimization inhttp://dx.doi.org/10.1016/j.asoc.2016.07.031
osition (Gbest—global search through neighboring particles expe-ience) during the search process to find the optimal solution withinhe search space of the problem.
The PSO algorithm has three steps: The algorithm process startshe initialization of randomly distributed populations within theearch solution space. Then at each iteration, the velocities and
377
378
etical RDN (b) before closing of tie switch (c) after closing of tie switch.
positions of each particle is updated using velocity and positionequations expressed as below,
Vt+1i
= ω.Vti + C1rand(1)(Pbestti − Xtid) + C2rand(2)(Gbest
ti − Xtid) (20)
Xt+1i
= (Xti
+ Vt+1i
), i = 1, 2, ......., N, d=1,2,......, D (21)
where “N” is the total number of particles; D is the dimension ofthe search space. Vt
isymbolize the velocity of the ith particle at
the tth iteration;Xtid
is the ith particle position at the tth iteration.The term “�” indicate the inertia weight which provides a balancebetween exploration and exploitation. The C1 and C2 are positivesocial weight parameters, which has control over the preeminenceand neighboring particle’s position on the current one. “Pbest” imi-tates the previous best position of the particle and “Gbest” knownas global best position which provides the best “Pbest” amongst allthe Pbest position; rand(1) and rand(2) are the random real value inthe range of [0,1].
ach of network reconfiguration with distributed generation radial distribution networks, Appl. Soft Comput. J. (2016),
Karaboga and Basturk [28] proposed artificial bee colony algo-rithm (ABC) from the foraging nature of honey bee swarm to solvethe constrained optimization problems. The ABC algorithm con-sists of the following groups; employed bee, onlooker bee and scout
379
380
381
382
ING ModelA
6 ompu
bssfsn
ieXtrf
V
wi
sb
Fs
F
ait
X
wf
nsu
roesatttiiiiivi
3
oaSes
for each fitnessj using Eq. (1).Step 5: Onlooker bee phase initiation: Based on the employed
bee phase best fitness probability, the new food source position
383
384
385
386
387
388
389
390
391
392
393
394
395
396
397
398
399
400
401
402
403
404
405
406
407
408
409
410
411
412
413
414
415
416
417
418
419
420
421
422
423
424
425
426
427
428
429
430
431
432
433
434
435
436
437
438
439
440
441
442
443
444
445
446
447
448
449
450
451
452
453
454
455
456
457
458
459
460
461
462
463
464
465
466
467
468
469
470
471
472
473
474
475
476
477
478
479
480
481
482
483
484
485
486
487
488
489
490
491
492
493
494
ARTICLESOC 3714 1–22
M. K., J. S. / Applied Soft C
ee. The role of employed bees is to search for the food source andhare the information to the onlooker bees; based on the better foodource information shared from the employed bees, the onlookerurther searches the better food source neighboring around theelected food source. If any food source is found abandoned, theew food source is randomly searched by the scout bees.
Let us suppose the solution space of the optimization problems D—dimensional. The algorithm is initiated with randomly gen-rated population and it can be represented as Xj = (Xj1, Xj2,. . ..jD) where j D {1,2,. . ..,SN). SN is the size of the food source. Then
he employed bee’s could modify the solution in its memory byandomly selecting a neighboring food source position using theollowing Eq. (22),
t+1j
= (Xtj + rand(1)(Xtj − Xtk) (22)
here K D {1,2,. . .. . .,SN) is chosen randomly with j /= K; rand (1)s a random number within the range of [−1,1].
In onlooker bee phase, the onlooker bees start searching theolution around the best food source provided by the employedees using Eq. (22) based on the probability value “Pj” as in Eq. (23)
Pj = fitnessj∑HMSj=1 fitnessj
(23)
itnessj is the fitness value of jth the solution. The fitness of eacholution is computed by Eq. (24) as,
itnessj = [1/(1 + min.fitness)] (24)
During this improvisation process, if any food source foundbandoned after a predetermined number of cycles (called as limit)t is replaced with new random food source by the scout bee usinghe following Eq. (25).
t+1j
= (Xmax + rand(0, 1)(Xmax − Xmin) (25)
here Xmax, Xmin are the maximum and minimum bound of theood source position.
If the new candidate food source position has better quality (fit-ess function) than the previous one, then such candidate foodource is retained by replacing the older one. Repeat the above stepsntil the optimal solution is obtained.
The standard ABC algorithm encounters the problem in explo-ation and exploitation ability when handling high dimensionalptimization problems. The solution search space is explored andxploited by the employee and onlooker bees using Eq. (22) isuperior in terms of tracing diverged solution (well explorationbility) but fails to exploit within the search space. Greedy selec-ion method on the bee’s trajectories will lead the ABC algorithmo trap into local optima. In this work, to cope with this problem,he intensification and diversification capability of ABC algorithms balanced by integrating the PSO into the ABC algorithm. Themproved ABC algorithm (PABC) updates the bee’s position by mak-ng use of Eq. (26). The inertia weight and acceleration constants arencorporated into the employed bee phase of the ABC algorithm tomprove the searching capability of the bees. These capabilities alle-iate the premature convergence and avoidance of the local optiman standard ABC algorithm.
.3. Proposed hybrid HS-PABC algorithm
It is a common practice to combine different search procedure tovercome the deficiencies of the individual heuristic optimization
Please cite this article in press as: M. K., J. S., Integrated approand shunt capacitors placement for power loss minimization inhttp://dx.doi.org/10.1016/j.asoc.2016.07.031
lgorithms in order to provide a robust optimization technique.tandard HSA is incompetent in local search ability [27] and itncounters the problem of slow convergence, traps in local optimalolution over multi model fitness landscape. The local searching
PRESSting xxx (2016) xxx–xxx
capability of the HSA has been enhanced by crossbreeding the HSAwith other optimization algorithms [22].
It is evident that in the HS algorithm the harmony memory (HM)plays a vital part which decides the algorithm performance. In thiswork, to augment the solution accuracy of the standard HSA, theparticle ABC algorithm is integrated and a new HS-PABC has beenproposed. In hybrid HS-PABC algorithm, the HMV in the HSA istaken as the initial food source by the PABC algorithm to producea diverse range of potentially optimal solutions while at the sameinstant it intensify its search in the vicinity of optimal solution. Newimprovised HMV values are used as initial values by the PABC algo-rithm to search for optimal solution. If the locally optimized vectorproduced by the PABC algorithm has better fitness value than thosein the HMV, they will be retained in the HM. The harmony vec-tor (HMV) in the HSA is improved by utilizing the potency of bothheuristic algorithms to get the best solution. Besides this hybridalgorithm requires the tuning of the control parameters such asHarmony Memory Size (HMS), Pitch Adjustment Rate (PAR), Har-mony Memory Consideration Rate (HMCR) and Maximum CycleNumber (MCN). The flowchart of the proposed hybrid HS- PABCalgorithm is presented in Fig. 2.
The computational procedure of the proposed hybrid HS-PABCalgorithm is detailed as follows,
Step 1: Initialize the parameters of HS-PABC algorithm; HMS,PAR, HMCR, MCN.
Step 2: Initialize the Harmony Memory Vector (HMV) with ran-dom population
Step 3: Cycle = Cycle + 1Step 4: Employed bee phase initiation: Random populations
generated in the HMV is improvised by using Eq. (26) to updatethe bee’s position.
Xjknew = [�Xjk + C1�jk(Xbest
k − Xjk) + C2�jk(Xnk − Xjk)] (26)
where Xjknew is the new feasible solution on its preceding solution
Xjk. The term “�” indicate the inertia weight utilized to modify thecurrent position of bees with respect to its prior one. To exploreglobal search at initial stage of the search, the inertia weight isdeclined linearly from 0.9 to 0.4 as suggested in [51] to exploreglobal search initially. The C1 and C2 are positive social weightparameter values chosen as 2.05 [51]. Xbest reflects the finest foodsource amongst the population. Xnk is the random neighbor valueof the population and �jk is the random value in the range of[0,1]. After the estimate of each candidate food source position, thegreedy selection procedure is applied among the solution obtainedthrough step 2 and 4. The onlooker bees utilize the nectar infor-mation shared from the employed bees and select the food sourcebased on the probability value accompanied with that food sourcePj from Eq. (27).
Pj = fitnessj∑HMSj=1 fitnessj
(27)
The fitness value of jth solution is taken as fitnessj. The fitnessof each solution is computed by Eq. (28) as,
Fitnessj = [1/(1 + min.fitness)] (28)
“min.fitness” is obtained through the BFS based power flow solution
ach of network reconfiguration with distributed generation radial distribution networks, Appl. Soft Comput. J. (2016),
Xjknew in the vicinity of food source Xj,k
old is determined by usingEq. (29)
Xjknew = {Xj,k
old + u(Xj,kold − Xn,k)} (29)
495
496
497
ARTICLE IN PRESSG ModelASOC 3714 1–22
M. K., J. S. / Applied Soft Computing xxx (2016) xxx–xxx 7
sed hy
Dwrlv
498
499
500
501
502
503
504505
Fig. 2. Flowchart of propo
where n D {1,2,. . ..CS} and k D {1,2,. . .. D} are random indices; is the total number of parameters in the optimization problem
Please cite this article in press as: M. K., J. S., Integrated approand shunt capacitors placement for power loss minimization inhttp://dx.doi.org/10.1016/j.asoc.2016.07.031
hich decides the dimension of the problem; n /= j; ‘u’ signifies aandom number produced which lies between [−1,1]. If the calcu-ated value of the candidate food position Xjk
new using Eq. (25) mayiolate the predefined boundary, resetting scheme is employed,
brid HS-PABC algorithm.
which put the violating boundary values to the adjoining boundarylimit value, i.e.,
ach of network reconfiguration with distributed generation radial distribution networks, Appl. Soft Comput. J. (2016),
Xjknew = Xjmin, ifXjk
new ≤ Xjmin
Xjknew = Xjmax, ifXjk
new ≥ Xjmax
506
507
ING ModelA
8 ompu
g
pt
X
Hr
e
o
C
3
m(
Scenario 2: Optimal allocation and sizing of DG unitsScenario 3: Simultaneous allocation (location and sizing) of DG
units and shunt capacitors
508
509
510
511
512
513
514
515
516
517
518
519
520
521
522
523
524
525
526
527
528
529
530
531
532
533
534
535
536
537
538
539
540
541
542
543
544
545
546
547
548
549
550
551
552
553
554
555
556
557
ARTICLESOC 3714 1–22
M. K., J. S. / Applied Soft C
After the estimation of each candidate food source position, thereedy selection method is adopted.
Step 6: Scout bee phase initiation: If any abandoned solution isresent, it is swapped with a new arbitrary solution generated byhe scout bee phase using Eq. (30).
jknew = min(Xk
j) + rand(0, 1)[max(Xkj) − min(Xk
j)] (30)
Step 7: Retain the best solution obtained so far in HMVStep 8: Improvise the populations (decision variables) in the
MV by HMCR, PAR and Random selection steps in the HS algo-ithm.
Step 9: Evaluate new HMV and check new HMV is better thanxisting HMV. If so, update HMV
Step 10: Cycle = Cycle + 1Step 11: Check if the stopping criterion is met. i.e., total number
f MCN is reached.Step 12: If not, go to step 4 and reiterate from step 4 until
ycle = MCN
.4. The pseudo code of proposed hybrid HS-PABC algorithm
Step 1: Initialize the parameters: Machine cycle (MCN), Har-ony memory size (HMS), Harmony memory consideration rate
HMCR) and Pitch Adjustment Rate (PAR)Step 2: Initialize HMS.For i = 1 to HMS doFor j = 1 to D do
Xij = lbj + rand ()*(ubj − lbj)Step 2.1: Calculate fitness value.
F(xi) = Evaluate(xi)Step 3: Start Cycle = 1
Please cite this article in press as: M. K., J. S., Integrated approand shunt capacitors placement for power loss minimization inhttp://dx.doi.org/10.1016/j.asoc.2016.07.031
PRESSting xxx (2016) xxx–xxx
Step 6: IF f(Xjnew) > f(Xj), update the HM as Xj = Xi
new
Step 7: Retain the best solution in the HMV.Step 8: Next Cycle = Cycle + 1Step 9: IF Cycle number < MCN, goto step 4; else stop the itera-
tive procedure
4. Numerical results and discussions
This article proposes an efficient hybrid heuristic search algo-rithm for the enhancement of distribution network performance byimplementing the four test scenarios with an objective of optimiz-ing the multi objective performance improvement indices such aspower loss reduction, bus voltage deviation and branch current ofthe RDN. The backward/forward sweep based power flow method(BFS) is utilized to find the load flow solution [50]. The proposedhybrid optimization algorithm is applied on 69 and 118 node radialtest networks on a Matlab7.7.0 platform. For both the test net-works, substation voltage is taken as 1.0 p.u. The four test scenariosat seven discrete load levels are,
Scenario 1: Optimal allocation (location and sizing) of shuntcapacitors
ach of network reconfiguration with distributed generation radial distribution networks, Appl. Soft Comput. J. (2016),
ARTICLE IN PRESSG ModelASOC 3714 1–22
M. K., J. S. / Applied Soft Computing xxx (2016) xxx–xxx 9
Table 1Average load duration schedule of a day.Q8
Time duration % load level No of hours/day No of hours/year
6 a.m.–7 a.m. 40 1 3657 a.m.–8 a.m. 50 1 3658 a.m.–9 a.m. 60 1 3659 a.m.–1 p.m. 80 4 14601p.m.–2 p.m. 70 1 3652 p.m.–5 p.m. 100 3 10955 p.m.–6 p.m. 90 1 3656 p.m.–9 p.m. 80 3 10959 p.m.–12 a.m. 60 3 1095
a
itsviiTDTt
4
rcr7bfiwb6
rThiirl6(6tbicuaatlolaa
resu
lts
of
69
nod
e
RD
N
at
dif
fere
nt
load
leve
ls
obta
ined
wit
h
hyb
rid
HS-
PAB
C.
el
Fixe
d/S
wit
ched
Cap
acit
orSc
enar
io-I
Scen
ario
-II
Scen
ario
-III
Scen
ario
-IV
Cap
acit
or(k
VA
R)
@n
ode
(67)
Cap
acit
or(k
VA
R)
@n
ode
(23)
Cap
acit
or(k
VA
R)
@n
ode
(61)
DG
1(k
W)
@
nod
e(6
3)D
G2
(kW
)
@
nod
e(6
1)D
G3
(kW
)
@
nod
e(6
4)D
G
(kW
)
@
nod
e(6
0)C
apac
itor
(kV
AR
)
@n
ode
(19)
DG
(kW
)
@
nod
e(6
0)C
apac
itor
(kV
AR
)
@n
ode
(19)
Op
tim
alto
pol
ogy
(op
ensw
itch
es)
Fixe
d
200
100
500
9.44
248.
51
122.
03
379.
57
200
343.
04
250
10-1
2-70
-61
-58
Swit
ched
50
–
150
111.
46
263.
02
100.
50
473.
24
50
433.
45
200
10-1
2-70
-62
-57
Swit
ched
50
50
100
190.
77
197.
45
181.
76
563.
37
100
527.
75
50
69-1
2-70
-61
-57
Swit
ched
100
–
150
118.
06
341.
88
205.
05
661.
01
50
612.
35
100
69-1
2-70
-61
-56
Swit
ched
–
50
350
361.
08
163.
89
235.
01
747.
37
50
706.
12
100
69-1
2-70
-61
-57
Swit
ched
50
50
–
327.
99
286.
48
241.
47
842.
34
50
819.
46
150
69-1
2-70
-61
-58
Swit
ched
–
–
50
422.
04
286.
48
241.
47
948.
44
–
897.
14
200
10-1
2-70
-61
-55
558
559
560
561
562
563
564
565
566
567
568
569
570
571
572
573
574
575
576
577
578
579
580
581
582
583
584
585
586
587
588
589
590
591
592
593
594
595
596
597
598
599
600
601
602
603
604
605
606
12 a.m.–6 a.m. 40 6 2190Total annual hours 8760
Scenario 4: Simultaneous optimal network reconfigurationlong with DG units and shunt capacitor compensation
The daily load schedule for the RDN is modeled for 24 h timenterval of the day and it is assumed to be repetitive for 365 dayso get the annual load schedule. The average daily load intervalchedule is depicted in Table 1. There are seven discrete load levelsarying from 40% to 100% as shown in Fig. 3. The DG unit’s injections considered as a negative PQ load with real and reactive powernjection capability with 0.85 lagging as an operating power factor.he status of sectionalizing and tie switches, location and size ofG unit and shunt capacitors are taken as the decision variables.he maximum penetration limit of DG unit is limited to 50% of theotal system real power demand.
.1. Simulation results of 69 node RDN
To test the effectiveness of the proposed hybrid HS-PABC algo-ithm a medium scale 69 node radial distribution network isonsidered with the real power load demand of 3802.1 kW andeactive power load demand of 2694.5 kVAR. This test system has3 line segments with five tie switches. The tie switches are num-ered from 69 to 73 and the sectionalizing switches are numberedrom 1 to 68.The active and reactive power loss at nominal load levels 224.97 kW and 102.15 kVAR respectively. The system operates
ith the nominal bus voltage magnitude of 12.66 kV with 100 MVAase. Minimum bus voltage magnitude of 0.9092 p.u. is observed at5th node. The test network data are obtained from [7].
The simulation results obtained with proposed hybrid algo-ithm and standard HSA for scenarios I–IV is summarized inables 2 and 3. The real power loss reduction obtained with HSA andybrid HS-PABC algorithm is summarized in Table 4. From Table 4,
t is observed that at 40% load level the base case loss of 32.50 kWs reduced to 21.65 kW, 10.54 kW, 11.09 kW and 7.07 kW withespect to the scenarios I–IV using hybrid approach. At mediumoad level (70%), the power loss reduction is from 104.52 kW to8.91 kW, 32.99 kW, 35.37 kW and 21.72 kW. At nominal load level100%) the power loss reduction is from 224.96 kW to 145.68 kW,9.08 kW, 73.11 kW and 48.62 kW. The bus voltage profile withhe scenarios I–IV at 100% load level is depicted in Fig. 4. It cane seen that, the proposed hybrid algorithm provides a significant
mprovement of minimum bus voltage profile with scenario-IV asompared with other scenarios. Fig. 5 depicts the optimal reconfig-ration topology and directed graph representation of 69 node RDNlong with DG units and shunt capacitor placement (Scenario-IV)t 100% load level. It is observed that scenario-IV provides bet-er loss reduction with improved bus voltage profile at all loadevel as compared with scenario-III. Since the power flow path is
Please cite this article in press as: M. K., J. S., Integrated approach of network reconfiguration with distributed generationand shunt capacitors placement for power loss minimization in radial distribution networks, Appl. Soft Comput. J. (2016),http://dx.doi.org/10.1016/j.asoc.2016.07.031
ptimized in scenario IV by retaining the DG and shunt capacitorocation obtained in scenario-III (@ node 19 for capacitor locationnd @ node 60 for DG unit location) better power loss reduction ischieved. Table 5 shows the minimum bus voltage profile obtained Ta
ble
2Si
mu
lati
on
%Lo
ad
lev
40
50
60
70
80
90
100
ARTICLE IN PRESSG ModelASOC 3714 1–22
10 M. K., J. S. / Applied Soft Computing xxx (2016) xxx–xxx
Fig. 3. Daily load schedule.
0.9
0.92
0.94
0.96
0.98
1
1 6 11 16 21 26 31 36 41 46 51 56 61 66
Bus V
olta
ge (p
.u)
Bus Number
Base Case Ca pac itor Placeme ntDG Uni ts PlacementSimulta neous DG un its + Capa cito r Placeme ntSimulta neous recon f + DG un its + capa citors
Fig. 4. Bus voltage profile of 69 node RDN with different scenarios with hybrid HS-PABC algorithm at 100% load level.
wsav
niaFe
4
st1ltrn(s
607
608
609
610
611
612
613
614
615
616
617
618
619
620
621
622
623
624
625
626
627
628
629
630
631
632
633
634
635
636
637
638
639
640
641
642
643
644
645
646
ith standard HSA and hybrid HS-PABC algorithm. The results pre-ented in Tables 4 and 5 reflect the efficiency of the proposed hybridlgorithm over standard HSA with respect to loss reduction andoltage profile improvement in all load levels.
The active power loss in each branch of the RDN with the sce-arios I–IV at 100% load level is depicted in Fig. 6. The reduction
n total real power loss of the RDN with scenario-IV at 40%, 70%nd 100% load level using hybrid HS-PABC algorithm is depicted inig. 7. It is revealed that scenario-IV offers more loss reduction withnhanced bus voltage profile as compared with other scenarios.
.2. Simulation results of 118 node RDN
To show the suitability of the proposed hybrid approach on largecale RDN it is applied on 118 node balanced three phase radial dis-ribution network. The system operates under 11 kV voltage level at00 MVA base with active load demand of 22709.7 kW and reactive
oad demand of 17041.1 kVAR respectively. The active and reac-ive power loss before compensation is 1298.1 kW and 978.74 kVAR
Please cite this article in press as: M. K., J. S., Integrated approand shunt capacitors placement for power loss minimization inhttp://dx.doi.org/10.1016/j.asoc.2016.07.031
espectively. The line and load data are obtained from [8]. This testetwork consists of 118 nodes with 117 sectionalizing switchesnormally closed) and fifteen tie switches (normally open). The tiewitches are numbered from 118 to 132.
Simulation results obtained using HS and hybrid HS-PABC algo-rithm at different load level of the network along with the candidatenodes for the placement of DG units and shunt capacitors are pre-sented in Tables 6–11. The real power loss reduction obtainedwith HSA and hybrid HS-PABC algorithms with four test scenar-ios are summarized in Table 12. From Table 12, it is observedthat loss reduction realized using proposed hybrid algorithm atlight load level (40%) is from 187.17 kW (base case) to 126.13 kW,59.59 kW, 66.75 kW and 54.91 kW with respect to the scenariosI–IV. At medium load level (70%), the power loss reduction is from602.18 kW to 399.78 kW, 185.32 kW, 209.21 kW and 163.91 kWwith respect to the scenarios I–IV. At nominal load level (100%) thepower loss reduction is from 1298.09 kW to 847.75 kW, 384.65 kW,435.62 kW and 340.12 kW with respect to the scenarios I–IV. Itis observed that scenario-IV provides better loss reduction withimproved bus voltage profile at all load level as compared withother scenarios. Fig. 8 depicts the optimal reconfiguration topol-ogy and directed graph representation of 118 node RDN along withDG units and shunt capacitor placement (Scenario-IV) at 100% load
ach of network reconfiguration with distributed generation radial distribution networks, Appl. Soft Comput. J. (2016),
level.Table 13 shows the minimum bus voltage profile with differ-
ent scenarios obtained with standard HSA and hybrid HS-PABCalgorithms. It can be inferred from the results presented in
647
648
649
650
ARTICLE IN PRESSG ModelASOC 3714 1–22
M. K., J. S. / Applied Soft Computing xxx (2016) xxx–xxx 11
Fig. 5. Directed graph representation of optimal topology of 69 Node RDN (Scenario-IV at 100% load level).
Ttwbdpillth
651
652
653
654
655
656
657
658
659
660
661
662
663
664
665
666
ables 12 and 13, that the proposed hybrid algorithm provides bet-er loss reduction with improved bus voltage profile as comparedith the standard HSA (with scenarios I–IV) at all load levels. The
us voltage profile with the scenarios I–IV at 100% load level isepicted in Fig. 9. It can be seen that, the proposed hybrid algorithmrovides a significant improvement of minimum bus voltage profile
n scenario-IV as compared with other scenarios. The active power
Please cite this article in press as: M. K., J. S., Integrated approand shunt capacitors placement for power loss minimization inhttp://dx.doi.org/10.1016/j.asoc.2016.07.031
oss in each branch of the RDN with the scenarios I–IV at 100% loadevel is depicted in Fig. 10.The reduction in total real power loss ofhe RDN with scenario-IV at 40%, 70% and 100% load levels usingybrid HS-PABC algorithm is depicted in Fig. 11. It can be inferred
that the combined approach of optimal network reconfigurationproblem along with DG units and shunt capacitors compensation(Scenario-IV) yields more loss reduction with improved bus voltageprofile as compared with other scenarios.
4.3. Parameter settings
ach of network reconfiguration with distributed generation radial distribution networks, Appl. Soft Comput. J. (2016),
An empirical study is performed to determine the impact ofcontrol parameters towards the solution quality of the hybrid HS-PABC algorithm. To demonstrate the effects of single parameterchange on the fitness function for 69-node RDN, 12 different cases
667
668
669
670
Please cite this article in press as: M. K., J. S., Integrated approach of network reconfiguration with distributed generationand shunt capacitors placement for power loss minimization in radial distribution networks, Appl. Soft Comput. J. (2016),http://dx.doi.org/10.1016/j.asoc.2016.07.031
ARTICLE IN PRESSG ModelASOC 3714 1–22
12 M. K., J. S. / Applied Soft Computing xxx (2016) xxx–xxx
Fig. 6. Variation of power loss at each line of 69 node RDN with different scenarios with hybrid HS-PABC algorithm at 100% load level.
0
50
100
150
200
250
40%70%100%
224.96
104.52
32.548.62
21.727.07
Pow
er l
oss
(k
W)
% load le vel
Base Case Scenario-I Scenario-II Scenario-III Scenario-IV
Fig. 7. Power loss reduction in 69 node RDN with different scenarios with hybrid HS-PABC algorithm at 100% load level.
Fig. 8. Directed graph representation of optimal reconfiguration topology of 118 node RDN (Scenario-IV at 100% load level).
ARTICLE IN PRESSG ModelASOC 3714 1–22
M. K., J. S. / Applied Soft Computing xxx (2016) xxx–xxx 13
0.85
0.9
0.95
1
1.05
1 6 11 16 21 26 31 36 41 46 51 56 61 66 71 76 81 86 91 96 101 106 111 116Bu
s Vol
tage
(p.u
)
Bus NumberBase Case Capacitor Placement DG units Placement Simultaneous DG units + capacitor placementSimultaneous reconf + DG units + capacitors placement
Fig. 9. Bus voltage profile of 118 node RDN with different scenarios (100% load level).
Fig. 10. Variation of real power loss at each line of 118 node RDN with different scenarios (100% load level).
0
500
1000
1500
%04%07%001
1298.09
602.18
187.17340.12
163.91 54.91
Po
wer
lo
ss (
kW
)
% Load level
Base Case Scenario-I Scenario-II Scenario-III Scenario-IV
Fig. 11. Real power loss reduction in 118 node RDN with different scenarios (100% load level).
aiiHpvhml(
the empirical study, the parameters chosen for 69 node network
671
672
673
674
675
676
677
678
679
680
681
682
683
684
685
re tested as shown in Table 14. Each case is tested over 50 runsn three stages. In stage 1–3, HMCR, HMS, PAR and MCN are var-ed respectively, and other parameters are kept constant. If theMCR value selection is too low, the algorithm behaves like aure random search which leads to slow convergence. If the HMCRalue is higher, all the harmonies in the HM are utilized, the other
Please cite this article in press as: M. K., J. S., Integrated approand shunt capacitors placement for power loss minimization inhttp://dx.doi.org/10.1016/j.asoc.2016.07.031
armonies are not well explored and this may lead to local opti-al solution. As shown in Table 14, large and small HMCR values
ead to a decrease in the solution quality. The pitch adjustmentPAR) is used to produce new solution around the existing best
solution. Selection of low PAR will limit the exploration and itcan lead to slow convergence. Higher PAR selection will lead thesolution to scatter around some potential optima similar to ran-dom search. The selection of larger and smaller values of HMS willdecrease the efficiency of the algorithm as seen in Table 14. From
ach of network reconfiguration with distributed generation radial distribution networks, Appl. Soft Comput. J. (2016),
are HMS = 20, HMCR = 0.9, PAR = 0.3, MCN = 15. Similarly for 118node RDN the control parameter values are selected as HMS = 50,MCN = 50, HMCR = 0.9, PAR = 0.3.
686
687
688
Please cite
this
article in
press
as: M
. K
., J.
S., In
tegrated ap
proach
of n
etwork
reconfi
guration
with
distribu
ted gen
erationan
d
shu
nt
cap
acitors
placem
ent
for
p
ower
loss
m
inim
ization
in
radial
d
istribution
n
etworks,
A
pp
l.
Soft
Com
pu
t.
J.
(2016),h
ttp://d
x.doi.org/10.1016/j.asoc.2016.07.031
AR
TIC
LE
IN P
RE
SS
G M
odelA
SOC
3714 1–22
14
M.
K.,
J. S.
/ A
pplied Soft
Computing
xxx (2016)
xxx–xxx
Table 3Simulation results of 69 node RDN at different load levels obtained with HSA.
%Load level Fixed/SwitchedCapacitor
Scenario-I Scenario-II Scenario-III Scenario-IV
Capacitor(kVAR) @node (67)
Capacitor(kVAR) @node (23)
Capacitor(kVAR) @node (61)
DG1 (kW)@ node (63)
DG2 (kW)@ node (61)
DG3 (kW)@ node (64)
DG (kW) @node (60)
Capacitor(kVAR) @node (19)
DG (kW) @node (60)
Capacitor(kVAR) @node (19)
Optimaltopology(openswitches)
40 Fixed 200 100 450 11.13 244.83 121 380.22 200 294.15 350 4-13-70-63-57
50 Switched 50 50 200 121.29 246 107.15 425.24 100 396 150 69-11-70-62-52
60 Switched 50 – 100 138.57 316.32 120.03 555.12 50 500.76 100 7-12-70-63-53
70 Switched 100 – 150 71.94 387.32 205.66 661.02 50 592.08 100 69-14-70-61-55
80 Switched 50 – 350 128.07 450.29 171.28 734.72 100 699.1 50 69-13-70-62-55
90 Switched – – – 133.97 503.13 207.74 823.75 – 791.7 100 7-11-70-62-58
100 Switched – – 100 104.06 538.30 227.03 900.02 – 878.9 300 69-12-70-63-57
Table 4Reduction in real power loss with different scenarios using HSA and hybrid HS – PABC.
%Load level Real Power loss (kW)
Before Compensation Scenario-ICapacitors @ node:(67),(23),(61)
Scenario-IIDG units @ node(63),(61),(64)
Scenario-IIIDG unit @ node(60)Capacitor @ node (19)
Scenario-IVReconfiguration alongwith DG @ node (60)Capacitor @ node (19)
HSA HS-PABC HSA HS-PABC HSA HS-PABC HSA HS-PABC
40 32.50 21.65 21.65 10.65 10.54 11.07 11.09 9.204 7.0750 51.59 34.41 34.26 16.64 16.61 19.92 17.53 13.16 11.4960 75.51 49.88 49.88 24.81 24.13 26.41 25.98 19.68 15.7170 104.52 68.91 68.91 32.99 32.99 35.37 35.37 23.57 21.7280 138.88 93.96 93.89 44.22 43.66 48.59 47.12 29.55 28.5090 178.92 117.44 117.58 56.23 55.41 61.52 59.98 43.87 36.48100 224.96 146.11 145.68 77.25 69.08 77.65 73.11 51.41 48.62
Please cite
this
article in
press
as: M
. K
., J.
S., In
tegrated ap
proach
of n
etwork
reconfi
guration
with
distribu
ted gen
erationan
d
shu
nt
cap
acitors
placem
ent
for
p
ower
loss
m
inim
ization
in
radial
d
istribution
n
etworks,
A
pp
l.
Soft
Com
pu
t.
J.
(2016),h
ttp://d
x.doi.org/10.1016/j.asoc.2016.07.031
AR
TIC
LE
IN P
RE
SS
G M
odelA
SOC
3714 1–22
M.
K.,
J. S.
/ A
pplied Soft
Computing
xxx (2016)
xxx–xxx
15
Table 5Minimum bus voltage magnitude with different scenarios using HSA and hybrid HS-PABC.
%Load level Minimum bus Voltage (p.u) @(Node)
Before compensation Scenario-I Scenario-II Scenario-III Scenario-IV
HSA @65 HS-PABC @ (65) HSA @ (65) HS-PABC @ (65) HSA @ (65) HS-PABC @ (65) HSA @63 HS-PABC @63
40 0.9656@18 0.9732 0.9739 0.9835 0.9836 0.9816 0.9815 0.9878 0.987850 0.9566@18 0.9675 0.9675 0.9792 0.9792 0.9750 0.9768 0.9849 0.986660 0.9474@18 0.9604 0.9604 0.9752 0.9755 0.9716 0.9720 0.9826 0.984870 0.9382@18 0.9539 0.9539 0.9711 0.9711 0.9673 0.9673 0.9816 0.982180 0.9287@18 0.9500 0.9500 0.9658 0.9670 0.9617 0.9620 0.9776 0.977990 0.9191@18 0.9409 0.9412 0.9616 0.9628 0.9563 0.9571 0.9738 0.9778100 0.9091@18 0.9330 0.9326 0.9538 0.9583 0.9504 0.9526 0.9733 0.9749
Table 6Optimal results obtained with hybrid HS-PABC algorithm—Scenario- I & II.
Load level Fixed/Switched Capacitor Scenario-I Scenario-II
Qc1 @(86) Qc2 @(82) Qc3 @(111) Qc4 @(49) Qc5 @(32) Qc6 @(107) Qc7 @(71) Qc8 @(96) DG1 @ (109) DG2 @(52) DG3 @(70)
40 Fixed 350 300 600 800 750 500 700 400 1315.43 1118.27 1195.4350 Switched 100 100 100 200 200 150 250 150 1646.62 1402.83 1498.1260 Switched 100 50 200 200 200 100 200 100 1978.76 1689.54 1802.4270 Switched 100 100 100 200 200 150 200 100 2311.87 1978.46 2108.3680 Switched 100 100 200 200 200 100 250 150 2646.42 2267.10 2415.1190 Switched 50 100 100 300 – 150 250 100 2980.41 2563.56 2726.37100 Switched – – 150 100 – – 200 50 3224.02 2757.83 2964.16
Please cite
this
article in
press
as: M
. K
., J.
S., In
tegrated ap
proach
of n
etwork
reconfi
guration
with
distribu
ted gen
erationan
d
shu
nt
cap
acitors
placem
ent
for
p
ower
loss
m
inim
ization
in
radial
d
istribution
n
etworks,
A
pp
l.
Soft
Com
pu
t.
J.
(2016),h
ttp://d
x.doi.org/10.1016/j.asoc.2016.07.031
AR
TIC
LE
IN P
RE
SS
G M
odelA
SOC
3714 1–22
16
M.
K.,
J. S.
/ A
pplied Soft
Computing
xxx (2016)
xxx–xxx
Table 7Optimal results obtained with hybrid HS-PABC algorithm—Scenario-III.
%Load level Fixed/Switched Capacitor Scenario-III
Qc1 Node (32) Qc2 Node (50) Qc3 Node (15) Qc4 Node (93) Qc5 Node (103) DG (kW)@ Node (71) DG (kW) @ Node (110)
40 Fixed 800 850 250 400 550 1147.65 1166.0450 Switched 200 250 100 50 150 1435.09 1458.1860 Switched 200 250 50 100 100 1727.78 1747.5670 Switched 250 300 50 100 150 2018.77 2041.4180 Switched 250 250 100 100 150 2315.09 2334.4790 Switched 250 250 50 100 150 2608.78 2629.85100 Switched 200 250 50 100 100 2904.64 2924.14
Table 8Optimal results obtained with hybrid HS-PABC algorithm—Scenario-IV.
%Load level Type of capacitor Scenario-IV
Fixed/Switched Capacitor size in kVAR Optimal DG units sizing in kW Optimal topology (Open Switches)
@ Node (32) @ Node (50) @ Node (15) @ Node (93) @ Node (103) @ Node (71) @ Node (110)
40 Fixed 750 750 300 250 550 1366.76 1594.45 25-23-39-42-34-60-51-125-126-96-127-107-79-130-12150 Switched 350 100 150 200 450 1527.37 1453.32 25-22-38-42-34-59-51-125-126-98-86-107-82-130-12160 Switched 250 200 – 150 – 1750.64 2090.32 25-22-38-42-34-61-50-125-126-97-87-131-80-130-12170 Switched 400 – 50 50 50 2252.65 2457.30 26-23-38-42-34-59-50-125-126-97-86-131-80-130-12180 Switched 200 300 150 – 100 2404.66 2769.04 26-22-39-42-34-59-50-125-126-97-86-131-80-130-12190 Switched – 350 – 150 250 2562.01 3146.19 25-23-39-42-34-60-51-125-126-97-86-131-80-130-121100 Switched – – 200 150 50 1366.76 1594.45 25-23-39-42-34-60-51-125-126-97-127-131-80-130-121
Please cite
this
article in
press
as: M
. K
., J.
S., In
tegrated ap
proach
of n
etwork
reconfi
guration
with
distribu
ted gen
erationan
d
shu
nt
cap
acitors
placem
ent
for
p
ower
loss
m
inim
ization
in
radial
d
istribution
n
etworks,
A
pp
l.
Soft
Com
pu
t.
J.
(2016),h
ttp://d
x.doi.org/10.1016/j.asoc.2016.07.031
AR
TIC
LE
IN P
RE
SS
G M
odelA
SOC
3714 1–22
M.
K.,
J. S.
/ A
pplied Soft
Computing
xxx (2016)
xxx–xxx
17
Table 9Optimal results obtained with HSA—Scenario-I & II.
Load level Fixed/Switched Capacitor Scenario-I Scenario-II
Qc1 @ (86) Qc2 @ (82) Qc3 @ (111) Qc4 @ (49) Qc5 @ (32) Qc6 @ (107) Qc7 @ (71) Qc8 @ (96) DG1 @ (109) DG2 @ (52) DG3 @ (70)
40 Fixed 300 300 650 1050 1000 450 750 350 1323.79 1276.49 1229.8250 Switched 100 50 150 50 550 150 250 200 1843.42 1654.55 1458.5060 Switched 150 150 100 200 50 150 150 100 2586.35 1686.08 1804.4970 Switched – 100 150 50 50 150 200 100 2515.81 1282.55 2108.8380 Switched – 150 150 100 50 150 250 150 2646.66 2269.03 2715.2990 Switched – 150 100 100 50 150 250 100 2998.59 2666.05 2821.19100 Switched – 700 50 100 50 150 100 – 3222.79 2759.82 2962.21
Table 10Optimal results obtained with HSA—Scenario-III.
%Load level Fixed/Switched Capacitor Scenario-III
Qc1 @Node (32) Qc2 @Node (50) Qc3 @Node (15) Qc4 @Node (93) Qc5 @Node (103) DG (kW) @ Node (71) DG (kW) @ Node (110)
40 Fixed 700 850 250 350 500 1143.54 1463.3350 Switched 250 250 100 150 150 1132.03 1758.5560 Switched 200 250 50 50 150 1525.89 1846.7370 Switched 200 300 50 50 150 1909.97 2231.7080 Switched 250 250 100 – 100 2119.18 2333.5690 Switched 250 250 50 250 100 2127.67 2632.98100 Switched 350 200 – 100 600 2936.47 3172.35
Please cite
this
article in
press
as: M
. K
., J.
S., In
tegrated ap
proach
of n
etwork
reconfi
guration
with
distribu
ted gen
erationan
d
shu
nt
cap
acitors
placem
ent
for
p
ower
loss
m
inim
ization
in
radial
d
istribution
n
etworks,
A
pp
l.
Soft
Com
pu
t.
J.
(2016),h
ttp://d
x.doi.org/10.1016/j.asoc.2016.07.031
AR
TIC
LE
IN P
RE
SS
G M
odelA
SOC
3714 1–22
18
M.
K.,
J. S.
/ A
pplied Soft
Computing
xxx (2016)
xxx–xxx
Table 11Optimal results obtained with HSA—Scenario-IV.
%Load level Type of capacitor Scenario-IV
Fixed/Switched Capacitor size in kVAR Optimal DG units sizingin kW Optimal topology (Open Switches)
@ Node (32) @ Node (50) @ Node (15) @ Node (93) @ Node (103) @ Node (71) @ Node (110)
40 Fixed 50 1000 100 550 350 1322.98 1226.79 119-20-124-42-34-61-52-125-126-96-86-131-82-130-12150 Switched 1700 550 300 150 250 1715.63 1540.30 25-22-124-42-33-58-50-125-126-98-86-131-81-130-12160 Switched 150 50 50 50 100 1911.94 2605.31 26-20-124-43-34-59-53-125-126-128-86-106-80-130-12170 Switched 50 50 50 50 50 1802.79 2662.57 26-23-39-44-33-59-50-125-126-97-87-104-81-130-12180 Switched – – 50 – – 2074.84 2930.43 25-22-38-43-33-59-50-125-126-97-74-131-80-130-12190 Switched 450 50 – – – 1795.53 3475.60 25-21-38-45-33-58-52-57-126-98-74- 131-79-130-121100 Switched 750 100 – – 600 2103.25 3984.31 25-21-39-44-34-59-52-125-70-128-127-131-80-130-121
Table 12Reduction in real power loss with scenarios: I–IV using HSA and hybrid HS-PABC.
%Load level Real Power loss (kW)
Before Compensation Scenario-ICapacitorsat node: (86),(82),(111),(49),(32),(107),(71),(96)
Scenario-IIDG units at node(109),(52),(70)
Scenario-IIIDG at node (71),(110)Capacitors at node(32),(50),(15),(93),(103)
Scenario-IVReconfiguration alongwith DG at node(71),(110)Capacitors at node(32),(50),(15),(93),(103)
HSA HS-PABC HSA HS-PABC HSA HS-PABC HSA HS-PABC
40 187.17 128.07 126.13 60.25 59.59 69.33 66.75 61.57 54.9150 297.14 200.12 199.52 96.15 93.59 110.95 105.03 99.53 86.7860 434.97 290.52 290.44 144.48 135.46 154.03 152.27 135.16 118.5170 602.18 400.06 399.78 197.31 185.32 210.83 209.21 179.83 163.9180 800.46 529.08 528.82 246.15 243.31 277.10 275.19 226.10 212.1190 1031.72 678.89 677.84 310.25 309.57 358.85 350.70 320.82 268.64100 1298.09 853.35 847.75 385.89 384.65 438.19 435.62 380.21 340.12
ARTICLE ING ModelASOC 3714 1–22
M. K., J. S. / Applied Soft Compu
Table 13Minimum bus voltage magnitude with scenarios: I–IV using HSA and hybrid HS-PABC.
%Load level Minimum bus Voltage (p.u) @(Node)
Beforecompensation@ (78)
Scenario-I Scenario-II
HSA @ (78) HS-PABC @ (78) HSA @ (47) H
40 0.9513 0.9658 0.9662 0.9844 050 0.9385 0.9581 0.9582 0.9803 060 0.9253 0.9487 0.9486 0.9763 070 0.9117 0.9396 0.9397 0.9713 080 0.8978 0.9312 0.9315 0.9682 090 0.8835 0.9226 0.9228 0.9641 0100 0.8687 0.9122 0.9127 0.9598 0
Table 14Impact of hybrid HS-PABC algorithm parameter variations on the fitness functionQ9(Scenario-IV) at 100% load level.
Stage Parameter settings Fitness
HMS HMCR PAR MCN
1 5 0.3 0.3 5 0.376405 0.3 0.3 10 0.354245 0.3 0.3 15 0.35324
2 5 0.3 0.3 15 0.353245 0.3 0.4 15 0.360585 0.3 0.5 15 0.35789
3 5 0.3 0.3 15 0.383245 0.7 0.3 15 0.357465 0.95 0.3 15 0.37668
4a
1
TS
TS
689
690
691
692
693
694
695
696
697
698
699
700
701
702
703
704
4 2 0.9 0.3 15 0.3775720 0.9 0.3 15 0.3440430 0.9 0.3 15 0.36862
.4. Robustness and convergence rate of hybrid HS-PABC and HS
Please cite this article in press as: M. K., J. S., Integrated approand shunt capacitors placement for power loss minimization inhttp://dx.doi.org/10.1016/j.asoc.2016.07.031
lgorithm
The formulated optimization problem (Scenarios-I–IV with00% load level) is solved 50 times repeatedly with random popula-
able 15tatistical study to test the robustness of the HS and hybrid HS −PABC algorithm with 50
Scenarios Max.fitness Min.fitness Mean fitness
HSA HS-PABC HSA HS-PABC HSA
I 0.525175272 0.515918944 0.515953 0.5158 0.517805366II 0.37432 0.370575 0.3703 0.36998 0.371701
III 0.381935 0.368598 0.3473269 0.347311 0.368063
IV 0.385383465 0.384051451 0.3457147 0.344045 0.368376043
able 16tatistical study to test the robustness of the HS and hybrid HS-PABC algorithm with 50 i
Scenarios Max.fitness Min.fitness Mean fitness
HSA HS-PABC HSA HS-PABC HSA
I 0.38792539 0.379718752 0.3811177 0.37960 0.38356404
II 0.186545315 0.180605777 0.180605 0.18060 0.183937826III 0.202517727 0.188078869 0.188329 0.188075 0.197148549IV 0.198168189 0.166000427 0.1630003 0.14670 0.180038661
PRESSting xxx (2016) xxx–xxx 19
Scenario-III Scenario-IV
S-PABC @ (47) HSA @ (55) HS-PABC HSA HS-PABC @ (51)
.9844 0.9821 0.9824@55 0.9857@97 0.9884
.9803 0.9784 0.9784@55 0.9846@51 0.9842
.9763 0.9744 0.9744@55 0.9796@43 0.9811
.9722 0.9713 0.9713@55 0.9774@51 0.9764
.9682 0.9674 0.9675@55 0.9732@51 0.9728
.9642 0.9635 0.9635@47 0.9629@58 0.9725
.9599 0.9589 0.9593@47 0.9655@ 99 0.9684
tion seeds. Both the algorithms are allowed to run over generationwith the same initial population. The robustness of hybrid approachover standard HSA is depicted in Tables 15 and 16. Scenarios-I–IVat 100% load level is considered to visualize the convergence plotof fitness function using HSA and hybrid HS-PABC algorithm asshown in Figs. 12 and 14(a–d). Undoubtedly Figs. 12 and 14(a–d)demonstrates the effectiveness of hybrid HS-PABC algorithm overHSA with stable and improved performance in terms of solu-tion accuracy and fast convergence to reach the optimal solutionin all the scenarios. This is because of high probability of bestsolutions updated in HMV by integrating PABC algorithm withHSA.
Table 17Statistical study using Wilcoxon signed rank sum test.
Algorithms Scenario 69 node RDN 118 node RDNP-Value P-Value
HS-PABC vsHSA
I 1.415656224849e-009 1.415656224849e-009II 1.415656224849e-009 1.390302548729e-009
ach of network reconfiguration with distributed generation radial distribution networks, Appl. Soft Comput. J. (2016),
independent runs at 100% load level – 69 node RDN.
Median Std deviation
HS-PABC HSA HS-PABC HSA HS-PABC
0.515901409 0.5176626 0.5158997 0.001751687 4.14644E-060.370158 0.371478 0.370096 0.000884 0.0001620.352237 0.36821 0.352933 0.008454 0.005464
0.353342188 0.370414 0.351459 0.010975562 0.009082648
ndependent runs at 100% load level –118 node RDN.
Median Std deviation
HS-PABC HSA HS-PABC HSA HS-PABC
0.379684188 0.3831836 0.3796802 0.001672306 9.68724E-06 0.18060575 0.1840858 0.1806057 0.001740765 7.13674E-09 0.18807596 0.1976716 0.18807568 0.003243615 7.62759E-07
0.153444577 0.1814853 0.1531285 0.009251209 0.004382506
III 1.795360263143e-007 1.415656224849e-009IV 1.651603235538e-005 1.991413222361e-009
Please cite this article in press as: M. K., J. S., Integrated approach of network reconfiguration with distributed generationand shunt capacitors placement for power loss minimization in radial distribution networks, Appl. Soft Comput. J. (2016),http://dx.doi.org/10.1016/j.asoc.2016.07.031
ARTICLE IN PRESSG ModelASOC 3714 1–22
20 M. K., J. S. / Applied Soft Computing xxx (2016) xxx–xxx
(a)
(b)
(c)
0 50 10 0 15 0 200 25 0 30 0 350 400 45 0 500 550 60 00.51 5
0.516
0.51 7
0.51 8
0.519
0.52
0.52 1
0.522
0.523
Iterations
Fitness
( O
bje
ctive function)
Conver gence plot of fitness fun ction with HSA and Hybrid HS -PABC algo rithms ( Scenario-I)
HSA- Cap Placenment
Hybrid HS -PABC -Cap Placement
0 50 100 150 200 250 300 350 400 450 500 550 6000.37
0.371
0.372
0.373
0.374
0.375
0.376
Iterations
Fitness
(O
bje
ctive functions)
Conver gence plot of fitness fun ction with HSA and Hybrid HS -PABC algo rithms ( Sce nario-II )
HSA- DG Placement
Hybrid HS -PABC- DG Placement
0 50 100 150 200 250 300 350 400 450 500 550 6000.346
0.348
0.35
0.352
0.354
0.356
0.358
0.36
Convergence plot of fitness function with HSA and Hybrid HS -PABC algo rithms ( Sce nario- III)
Fit
ness
( O
bje
cti
ve f
un
cti
on
)
Iterations
HSA- DG + Capacitor Place ment
Hybrid HS -PABC - DG + Capacitor Place ment
(d)
0 50 100 15 0 200 250 300 350 400 450 500 550 6000.34
0.34 5
0.35
0.35 5
0.36
0.36 5
0.37
0.37 5
0.38
0.38 5
0.39
Convergence plot of fitness function with HSA and Hy brid HS -PABC algorithms ( Scenario-IV)
Iterations
Fitness
( O
bje
cti
ve functi
on)
HSA- Rec on + DG+Capa citor placement
Hybrid HS -PABC- Recon+ DG+ Capacitor placement
Fig. 12. (a–d) Convergence plot of fitness function with HS and hybrid HS-PABCalgorithms (Scenario- I–IV)—100% load level-69 node RDN.
(a)
(b)
(c)
0 500 10 00 1500 2000 2500 3000 3500 400 0 45 00 50000.37 8
0.38
0.38 2
0.38 4
0.38 6
0.38 8
0.39
0.39 2
Iterations
Fit
nes
s (
Ob
ject
ive
fun
ctio
n)
Convergence plot of fitness function with HS and hybrid HS-PABC algorithms ( Scenario-I)
HSA- Capacitor placement
Hybrid HS-PABC- Capacitor Placement
0 50 0 1000 1500 200 0 250 0 300 0 350 0 40 00 45 00 50 000.18
0.18 2
0.18 4
0.18 6
0.18 8
0.19
0.19 2
0.19 4Convergence plot of fitness fun ction with HS and hybrid HS-PABC algo rithms ( Sce nario-II)
Iterations
Fit
ness
( o
bje
cti
ve f
un
cti
on
)
HSA - DG Placement
Hybrid HS- PABC- DG Placement
0 50 0 100 0 150 0 200 0 250 0 30 00 35 00 4000 4500 500 00.18
0.19
0.2
0.21
0.22
0.23
Iter ations
Fit
ness
( O
bje
cti
ve f
uncti
on)
Conver gence plot of fitness fun ction with HSA and hybrid HS- PABC algo rithms ( Sce nario-III )
Hybrid HS-PABC DG + Capacitor Placement
HSA- DG + Ca pacitor Placement
(d)
0 30 0 600 90 0 120 0 15 00 180 0 21 00 240 0 27 00 3000 330 0 3600 390 0 42 00 450 0 48 00 50 000.15
0.16
0.17
0.18
0.19
0.2
0.21
0.22
0.23
Iter ations
Fit
ness
( O
bje
cti
ve f
un
cti
on
)
Convergence plot of fitness fun ction with HS and hybrid HS-PABC algorithms (Sce nario-IV)
Hybr id HS-PABC- Recon+ DG+ Capacitors
HSA - Rec on+ DG+ Capacitors
Fig. 14. (a–d) Convergence plot of fitness function with HSA and Hybrid HS-PABCalgorithm (Scenario- I–IV)—100% load level-118 node RDN.
(a) (b)
0.51
0.512
0.514
0.516
0.518
0.52
0.522
0.524
0.526
HS-PABC HSA
Fitn
ess
Algori thm
Scenario-I -69 node RDN
0.367
0.368
0.369
0.37
0.371
0.372
0.373
0.374
0.375
HS-PABC HSA
Fit
nes
s
Algori thm
Scenario-II - 69 nod e RDN
(c) (d)
0.32
0.33
0.34
0.35
0.36
0.37
0.38
0.39
HS-PABC HSA
Fit
nes
s
Algorithm
Scenario-III - 69 node RDN
0.32
0.33
0.34
0.35
0.36
0.37
0.38
0.39
HS-PABC HSA
Fit
nes
s
Algorithm
Scenario-IV - 69 node RDN
Fig. 13. (a–d) Box plot for the results of HS-PABC and HS algorithms—Scenarios I–IV for 69 node RDN.
ARTICLE ING ModelASOC 3714 1–22
M. K., J. S. / Applied Soft Compu
(a) (b
0.3740.3760.378
0.380.3820.3840.3860.388
0.39
HS-PABC HSA
Fit
nes
s
Algorithm
Scenario-I 118 node RDN
(c ) (d
0.18
0.183
0.186
0.189
0.192
0.195
0.198
0.201
0.204
HS-PABC HSA
Fit
nes
s
Algori thm
Scenario-III 118 node RDN
and H
4
qH
4dsiqoHo
4ptsd
HP
HH
snmsHwwbewm
Q7
705
706
707
708
709
710
711
712
713
714
715
716
717
718
719
720
721
722
723
724
725
726
727
728
729
730
731
732
733
734
735
736
737
738
739
740
741
742
743
744
745
746
747
748
749
750
751
752
753
754
755
756
757
758
759
760
761
762
763
764
765
766
767
768
Fig. 15. (a–d) Box plot for the results of HS-PABC
.4.1. Statistical analysisThe box plot and Wilcoxon rank sum test is used to show the
uality of the solution obtained by hybrid HS-PABC with respect toSA.
.4.1.1. Box-plot. Repeatedly 50 runs are performed to visualize theistribution of the samples in HS-PABC and HS algorithms with fourcenarios in box plot format as shown in Figs. 13 and 15(a–d). Its observed that the distribution of data based on minimum, firstuartile, median, third quartile, and maximum are close to eachther in proposed hybrid HS-PABC algorithm when compared withSA. The results in box plots reveal the robustness and superiorityf the proposed hybrid HS-PABC over standard HS algorithm.
.4.1.2. Wilcoxon rank sum test. Wilcoxon rank sum test is a non-arametric test utilized to show the statistical comparison betweenhe proposed hybrid HS-PABC and HS algorithm. This test assumeselection of different random samples with continuous probabilityistribution. The following hypotheses are tested.
0. The probability distributions of fitness values obtained by HS-ABC and HSA are identical.
1. Distribution of the fitness values differ between HS-PABC andSA
Wilcoxon rank sum test results are termed p-values (observedignificance level). When p-values less than or equal to level of sig-ificance (� = 5%), we reject the null hypothesis. The approximationethod is used to calculate the p-values and the results are pre-
ented in Table 17 to illustrate the efficiency of proposed hybrid
Please cite this article in press as: M. K., J. S., Integrated approand shunt capacitors placement for power loss minimization inhttp://dx.doi.org/10.1016/j.asoc.2016.07.031
S-PABC algorithm over standard HS algorithm. For both the RDNsith scenario-IV, all the p-values are less than 0.05 (refer Table 17)hich signifies the choice of selecting alternative hypothesis. It can
e concluded that these test implies a significant statistical differ-nce between the HS-PABC and standard HS algorithm. Thereforee can come to the conclusion that the proposed HS-PABC canaintain superior performance than the standard HSA.
769
770
771
772
PRESSting xxx (2016) xxx–xxx 21
)
0.176
0.178
0.18
0.182
0.184
0.186
0.188
HS-PAB C HSA
Fit
nes
s
Algorithm
Scenario-II 118 node RDN
)
0.12
0.14
0.16
0.18
0.2
0.22
0.24
HS-PABC HSA
Fit
nes
s
Algori thm
Scenario-IV - 118 node RDN
S algorithms—Scenarios I–IV for 118 node RDN.
5. Conclusion
This article presents a hybrid HS-PABC algorithm basedapproach for network reconfiguration problem along with optimallocation and sizing of DG units and shunt capacitors for optimalplanning of distribution networks. Different test scenarios are sim-ulated on 69 and 118 node distribution network at varying loadconditions. The hybrid HS-PABC algorithm offers a promising andpreferable performance in terms of power loss reduction, improvedbus voltage profile than standard HSA at all discrete load levels.From the simulation results, it is inferred that integrated approachof network reconfiguration problem with the simultaneous instal-lation of DG units and shunt capacitors is an efficient approachtowards more loss reduction with the enhanced bus voltage profileas compared with individual loss reduction approaches. A statisticalanalysis is performed using 50 independent trials for each approachon the test system. To measure and prove the quality of the solu-tions obtained by proposed approach, box plot and Wilcoxon ranksum test are performed. The HS-PABC is a co-evolutionary algo-rithm that runs in parallel to accomplish the optimal solution inhigh-dimensional optimization problems. The simulation resultsconfirm the capability of proposed HS-PABC algorithm to reachthe optimal solution with faster convergence rate than standardHS algorithm. Finally, this work provides interesting basis forcombining promising aspects of different evolutionary algorithmtechniques into a new hybrid meta-heuristic algorithm that showssignificant performance.
References
[1] Leonardo W. De Oliveira, Edimar J. de Oliveira, Flávio V. Gomes, Ivo C. Silva,André L.M. Marcato, Paulo V.C. Resende, Artificial immune systems applied to
the reconfiguration of electrical power distribution networks for energy lossminimization, Int. J. Electr. Power Energy Syst. 56 (2014) 64–74.
[2] Almoataz Youssef Abdelaziz, F.M. Mohammed, S.F. Mekhamer, M.A.L. Badr,Distribution systems reconfiguration using a modified particle swarmoptimization algorithm, Electr. Power Syst. Res. 79 (11) (2009) 1521–1530.
[3] Saeed Teimourzadeh, Kazem Zare, Application of binary group search
ach of network reconfiguration with distributed generation radial distribution networks, Appl. Soft Comput. J. (2016),
optimization to distribution network reconfiguration, Int. J. Electr. PowerEnergy Syst. 62 (2014) 461–468.
[4] Anil Swarnkar, Nikhil Gupta, K.R. Niazi, Adapted ant colony optimization forefficient reconfiguration of balanced and unbalanced distribution systems forloss minimization, Swarm Evol. Comput. 1 (3) (2011) 129–137.
773
774
775
776
777
ING ModelA
2 ompu
[
[
[
[
[
[
[
[
[
[
[
[
[
[
[
[
[
[
[
[
[
[
[
[
[
[
[
[
[
[
[
[
[
[
[
[
[
[
[
[
[50] Teng Jen-Hao, A direct approach for distribution system load flow solutions,IEEE Trans. Power Deliv. 18 (3) (2003) 882–887.
[51] Leandro Dos Santos Coelho, Cezar Augusto Sierakowski, A software tool for
778
779
780
781
782
783
784
785
786
787
788
789
790
791
792
793
794
795
796
797
798
799
800
801
802
803
804
805
806
807
808
809
810
811
812
813
814
815
816
817
818
819
820
821
822
823
824
825
826
827
828
829
830
831
832
833
834
835
836
837
838
839
840
841
842
843
844
845
846
847
848
849
850
851
852
853
854
855
856
857
858
859
860
861
862
863
864
865
866
867
868
869
870
871
872
873
874
875
876
877
878
879
880
881
882
883
884
885
886
887
888
889
890
891
892
893
894
895
896
897
898
899
900
901
902
903
904
905
906
907
908
909
910
911
912
913
ARTICLESOC 3714 1–22
2 M. K., J. S. / Applied Soft C
[5] A. Mohamed Imran, M. Kowsalya, A new power system reconfigurationscheme for power loss minimization and voltage profile enhancement usingFireworks Algorithm, Int. J. Electr. Power Energy Syst. 62 (2014) 312–322.
[6] Rayapudi Srinivasa Rao, Sadhu Venkata Lakshmi Narasimham, ManyalaRamalinga Raju, Optimal network reconfiguration of large-scale distributionsystem using harmony search algorithm, IEEE Trans. Power Syst. 26 (3) (2011)1080–1088.
[7] J.S. Savier, Debapriya Das, Impact of network reconfiguration on lossallocation of radial distribution systems, IEEE Trans. Power Deliv. 22 (4)(2007) 2473–2480.
[8] D. Zhang, Z. Fu, L. Zhang, An improved TS algorithm for loss minimumreconfiguration in large scale distribution systems, Electr. Power Syst. Res. 77(2007) 685–694.
[9] Y. Mishima, K. Nara, T. Satoh, et al., Method for minimum-loss reconfigurationof distribution system by Tabu search, Electr. Eng. Jpn. 152 (2005) 18–25.
10] A.Y. Abedelaziz, F.M. Mohamed, S.F. Mekhamer, et al., Distribution systemreconfiguration using a modified Tabu search algorithm, Elect. Power Syst.Res. 80 (2010) 943–953.
11] Hamed Ahmadi, Jose R. Matri, Mathematical representation of radialityconstraints in distribution system reconfiguration problem, Int. J. Electr.Power Energy Syst. 64 (2015) 293–299.
12] R. Srinivasa Rao, Kumudhini Ravindra, K. Satish, S.V.L. Narasimham, Powerloss minimization in distribution system using network reconfiguration in thepresence of distributed generation, IEEE Trans. Power Syst. 28 (1) (2013)317–325.
13] A. Mohamed Imran, M. Kowsalya, D.P. Kothari, A novel integration techniquefor optimal network reconfiguration and distributed generation placement inpower distribution networks, Int. J. Electr. Power Energy Syst. 63 (2014)461–472.
14] Attia A. El-Fergany, Almoataz Y. Abdelaziz, Artificial bee colony algorithm toallocate fixed and switched static shunt capacitors in radial distributionnetworks, Electr. Power Compon. Syst. 42 (5) (2014) 427–438.
15] Pavlos S. Georgilakis, Nikos D. Hatziargyriou, Optimal distributed generationplacement in power distribution networks: models, methods, and futureresearch, IEEE Trans. Power Syst. 28 (3) (2013) 3420–3428.
16] Pavlos S. Georgilakis, Nikos D. Hatziargyriou, A review of power distributionplanning in the modern power systems era: models, methods and futureresearch, Electr. Power Syst. Res. 121 (2015) 89–100.
17] Duong Quoc Hung, Nadarajah Mithulananthan, R.C. Bansal, Analyticalexpressions for DG allocation in primary distribution networks, IEEE Trans.Energy Convers. 25 (3) (2010) 814–820.
18] S. Gopiya Naik, D.K. Khatod, M.P. Sharma, Optimal allocation of combined DGand capacitor for real power loss minimization in distribution networks, Int. J.Electr. Power Energy Syst. 53 (2013) 967–973.
19] Zong Woo Geem, Joong Hoon Kim, G.V. Loganathan, A new heuristicoptimization algorithm: harmony search, Simulation 76 (2) (2001) 60–68.
20] M. Mahdavi, M. Fesanghary, E. Damangir, An improved harmony searchalgorithm for solving optimization problems, Appl. Math. Comput. 188 (2007)1567–1579.
21] Kang Seok Lee, et al., The harmony search heuristic algorithm for discretestructural optimization, Eng. Optim. 37 (7) (2005) 663–684.
22] Bin Wu, Cunhua Qian, Weihong Ni, Shuhai Fan, Hybrid harmony search andartificial bee colony algorithm for global optimization problems, Comput.Math. Appl. 64 (8) (2012) 2621–2634.
23] K. Muthukumar, S. Jayalalitha, Optimal placement and sizing of distributedgenerators and shunt capacitors for power loss minimization in radialdistribution networks using hybrid heuristic search optimization technique,Int. J. Electr. Power Energy Syst. 78 (June) (2016) 299–319.
24] Z.W. Geem, Optimal cost design of water distribution networks usingharmony Search, Eng. Optim. 38 (2006) 259–277.
25] Z.W. Geem, Particle-swarm harmony search for water network design, Eng.Optim. 49 (2009) 297–311.
26] Pan Quan-Ke, P.N. Suganthan, M. Fatih Tasgetiren, A self-adaptive global bestharmony search algorithm for continuous optimization problems, Appl. Math.Comput. 216 (2010) 830–848.
27] M. Fesanghary, M. Mahdavi, M. Minary-Jolandan, Y. Alizadeh, Hybridizing
Please cite this article in press as: M. K., J. S., Integrated approand shunt capacitors placement for power loss minimization inhttp://dx.doi.org/10.1016/j.asoc.2016.07.031
sequential quadratic programming with HS algorithm for engineeringoptimization, Comput. Methods Appl. Mech. Eng. 197 (2008) 3080–3091.
28] D. Karaboga, B. Basturk, A powerful and efficient algorithm for numericalfunction optimization: artificial bee colony algorithm, J. Global Optim. 3(2007) 9459–9471.
PRESSting xxx (2016) xxx–xxx
29] D. Karaboga, B. Basturk, On the performance of artificial bee colony algorithm,Appl. Soft Comput. 8 (2008) 687–697.
30] D. Karaboga, Akay Bahriye, A comparative study of artificial bee colonyalgorithm, Appl. Math. Comput. 214 (2009) 108–132.
31] Z.W. Geem, Novel derivative of harmony search algorithm for discrete designvariables, Appl. Math. Comput. 199 (1) (2008) 223–230.
32] K.S. Lee, Z.W. Geem, A new meta-heuristic algorithm for continuousengineering optimization: harmony search theory and practice, Comput.Methods Appl. Mech. Eng. 194 (36–38) (2005) 3902–3933.
33] D. Manjarres, I. Landa-Torres, S. Gil-Lopez, J. Del Ser, M.N. Bilbao, S.Salcedo-Sanz, Z.W. Geem, A survey on applications of the harmony searchalgorithm, Eng. Appl. Artif. Intell. 26 (8) (2013) 1818–1831.
34] Swagatam Das, Arpan Mukhopadhyay, Anwit Roy, Ajith Abraham, Bijaya K.Panigrahi, Exploratory power of the harmony search algorithm: analysis andimprovements for global numerical optimization, IEEE Trans. Syst. ManCybern.—B: Cybern. 41 (1) (2011) 89–106.
35] Z.W. Geem, Optimal scheduling of multiple dam system using harmonysearch algorithm, Lect. Notes Comput. Sci. 4507 (2007) 316–323.
36] I. Ahmad, M. Gh. Mohammad, A.A. Salman, S.A. Hamdan, Broadcastscheduling in packet radio networks using harmony search algorithm, ExpertSyst. Appl. 39 (1) (2012) 1526–1535.
37] I. Landa-Torres, E.G. Ortiz-Garcia, S. Salcedo-Sanz, M.J. Segovia-Vargas, S.Gil-L’opez, M. Miranda, J.M. Leiva-Murillo, J. Del Ser, Evaluating theinternationalization success of companies through a hybrid groupingharmony search—extreme learning machine approach, IEEE J. Sel. Top. SignalProcess. 6 (4) (2012) 388–397.
38] Z.W. Geem, K.S. Lee, Y. Park, Application of harmony search to vehicle routing,Am. J. Appl. Sci. 2 (12) (2005) 1552–1557.
39] I. Landa-Torres, S. Gil-Lopez, S. Salcedo-Sanz, J. Del Ser, J.A. Portilla-Figueras, Anovel grouping harmony search algorithm for the multiple-type access nodelocation problem, Expert Syst. Appl. 39 (5) (2012) 5262–5270.
40] I. Landa-Torres, D. Manjarres, S. Salcedo-Sanz, J. Del Ser, S. Gil-Lopez, Amulti-objective grouping harmony search algorithm for the optimaldistribution of 24-h medical emergency units, Expert Syst. Appl. 40 (6) (2013)2343–2349.
41] S. Salcedo-Sanz, D. Manjarres, A. Pastor-Sanchez, J. Del Ser, J.A.Portilla-Figueras, S. Gil-Lopez, One-way urban traffic reconfiguration using amulti-objective harmony search approach, Expert Syst. Appl. 40 (9) (2013)3341–3350.
42] Chandan Kumar Shiva, V. Mukherjee, A novel quasi-oppositional harmonysearch algorithm for automatic generation control of power system, Appl. SoftComput. 35 (2015) 749–765.
43] Hai-bin Ouyang, Li-qun Gao, Xiang-yong Kong, Steven Li, De-xuan Zou,Hybrid harmony search particle swarm optimization with global dimensionselection, Inf. Sci. 346–347 (2016) 318–337.
44] Fuqing Zhao, Yang Liu, Chuck Zhang, Junbiao Wang, A self-adaptive harmonyPSO search algorithm and its performance analysis, Expert Syst. Appl. 42(2015) 7436–7455.
45] Bahriye Akay, Dervis Karaboga, Artificial bee colony algorithm for large-scaleproblems and engineering design optimization, J. Intell. Manuf. 23 (2012)1001–1014.
46] Dervis Karaboga, Bahriye Akay, A modified artificial bee colony (ABC)algorithm for constrained optimization problems, Appl. Soft Comput 11(2011) 3021–3031.
47] G. Li, P. Niu, X. Xiao, Development and investigation of efficient artificial beecolony algorithm for numerical function optimization, Appl. Soft Comput. 12(2012) 320–332.
48] Ahmad Nickabadi, Mohammad Mehdi Ebadzadeh, Reza Safabakhsh, A novelparticle swarm optimization algorithm with adaptive inertia weight, Appl.Soft Comput. 11 (2011) 3658–3670.
49] Zhiyong Li, Weiyou Wang, Yanyan Yan, Zheng Li, PS–ABC: a hybrid algorithmbased on particle swarm and artificial bee colony for high-dimensionaloptimization problems, Expert Syst. Appl. 42 (2015) 8881–8895.
ach of network reconfiguration with distributed generation radial distribution networks, Appl. Soft Comput. J. (2016),
teaching of particle swarm optimization fundamentals, Adv. Eng. Softw. 39(11) (2008) 877–887.
914
915