Materials Today: Proceedings xxx (xxxx) xxx
Contents lists available at ScienceDirect
Materials Today: Proceedings
journal homepage: www.elsevier.com/locate/matpr
Maximization of economy in distribution networks with most favorable
placement of distributed generators along with reorganization using
hybrid optimization algorithm
Vempalle Rafi ⇑, P.K. Dhal
Department of Electrical and Electronics Engineering, Vel Tech Rangarajan Dr. Sagunthala R&D Institute of Science and Technology, Chennai, India
a r t i c l e
i n f o
Article history:
Received 6 October 2020
Accepted 26 October 2020
Available online xxxx
Keywords:
Distributed generators
Hybrid optimization
Particle swarm optimization (PSO)
Radial distribution system
Salp Swarm Algorithm (SSA)
a b s t r a c t
Hybrid optimization is used to maximize the savings through reconfiguration, type-1 and type-2 distributed generators (DG) placement. The PSO and SSA are used for reconfiguration and DG placement
respectively. The major objective for this algorithm is to maximize the economy by minimizing the loss,
by including loss cost. The suggested algorithm is checked on two IEEE radial distribution system (RDS)
which are IEEE 119 RDS and IEEE 136 radial distribution network. The installation cost and maintenance
of DG is included in objective of hybrid optimization while installation of DG along with reconfiguration
and without reconfiguration.
Ó 2020 Elsevier Ltd. All rights reserved.
Selection and peer-review under responsibility of the scientific committee of the Emerging Trends in
Materials Science, Technology and Engineering.
1. Introduction
Utilization of generated power is at the generation side in the
reconfigured power system. The generators are called distributed
generators, which are installed at distribution side. The generators
(DGs). DGs are categorized to three types which compensates type
1, type II and type III, compensated watt power (kW), wattles
power (kVar) and apparent power (kVA) respectively. The distribution network is studied by wield load flow investigation. The radial
character of the network is not suitable to use, commercial load
flow techniques like Gauss-seidel method, Newton Raphson
method which are called commercial load flows. So, the RDS is analyzed by wield forward/ backward sweep load flow. The major
application of load flows is determining voltage profile, losses
and branch currents of the system. The losses of the network are
affecting the total savings of the system as considered the cost
for them. The losses of the system can be reduced by using compensated devices like static capacitor, DGs, D-STATCOM etc. The
system losses are minimized by compensating the installation
and maintenance costs for the systems. The system losses also
minimized by reconfiguration, devoid of any expense. Reorganize
problem, initially suggested by Merlin et al. [1]. It’s very prolonged
approach as probable device layout, where line parts situated fur⇑ Corresponding author.
nished with toggles. An algorithm established on the heuristic
guidelines and stupid multi-target technique to optimize the network [2]. Major consequence in the proposed algorithm is for the
selection of membership functions of the targets are not reported.
Genetic algorithm (GA) exterminates the problem with minimum
loss organization for distribution installation [3]. An algorithm to
resolve reconfiguration of network issue with optimal combinations of switches combinedly in the network was proposed to
diminish power losses [4]. The placement of capacitor to reorganize RDS was proposed using PSO [5]. In addition, Harmony Search
Algorithm (HSA) was proposed to address system network restructuring allocation crises in the event of DG [6]. Survey on unified
tracking and improvement voltage control calculations has been
introduced in [7]. GA and enhanced joined strategy for switchtrade in a DNR issue was used to accomplish a base misfortune
within the sight of DGs [8].The comparative study of DGs is proposed in [9]. The real power compensation DGs are designed with
PV system which are designed with inverter [10]. The preferable
placement and accurate size of DG in RDS are critical requirements
for the minimization of losses in the system [11]. The increment of
losses in system will influence the bus voltage, which motivates
the usage of voltage regulation through energy storage system
[12] Table 1.
The literature detailed up to now is illustrated only one heuristic optimization schemes for reconfiguration with preferable place-
https://doi.org/10.1016/j.matpr.2020.10.776
2214-7853/Ó 2020 Elsevier Ltd. All rights reserved.
Selection and peer-review under responsibility of the scientific committee of the Emerging Trends in Materials Science, Technology and Engineering.
Please cite this article as: V. Rafi and P.K. Dhal, Maximization of economy in distribution networks with most favorable placement of distributed generators
along with reorganization using hybrid optimization algorithm, Materials Today: Proceedings, https://doi.org/10.1016/j.matpr.2020.10.776
V. Rafi and P.K. Dhal
Materials Today: Proceedings xxx (xxxx) xxx
2.1. Development of BIBC matrix
Table 1
Algorithm parameters and their values.
PSO
The current equations can be found by wield KCL for Fig. 1 is:
SSA
Number of particles
Number of iterations
Max inertia
Min inertia
Intial velocity
Final velocity
30
50
0.9
0.2
2
2
Number of Searching Agents
Number of iterations
–
–
–
40
50
–
–
–
–
ment and accurate size of DG. In this study different optimization
algorithms such as PSO for reconfiguration, SSA for preferable position and size of the DG. The paper is divided in to four parts, such
as second part load flow, part three distribution system reorganization is discussed and part four comparative analysis of different
optimization techniques is proposed and part five results are
conclusions.
IB5 ¼ IL6
ð1Þ
IB4 ¼ IL5
ð2Þ
IB3 ¼ IL4 þ IL5
ð3Þ
IB2 ¼ IL3 þ IL4 þ IL5 þ IL6
ð4Þ
IB1 ¼ IL2 þ IL3 þ IL4 þ IL5 þ IL6
ð5Þ
Thus, the correlation among load currents & line currents in
matrix structure as:
½IB ¼ ½BIBC ½IL ð6Þ
The receiving end voltages can be calculated by forward sweep
as:
V q ðkÞ ¼ V p ðkÞ IB ðkÞ Z B ðkÞ
2. Methodology
For any study in the power system the load flows are the primary phase. The study like power quality, voltage quality, conductor thermal constraints, watt and reactive power losses. The load
flow analysis during this study is employed a well-known algorithmic program known as forward/ backward sweep algorithm. The
algorithmic program, load flow is started with assumption of flat
voltage profile. The currents are determined from the load by using
branch impedances and flat voltage profile in reverse direction, the
voltage of the buses is in forward way and this algorithm is called
forward / backward sweep algorithm. RDS reconfiguration hinge
on the number of loops to the system. The system loops are displaced by opening a branch called tie-line switch and the branches
of the system which are not set the loop called main line switch or
sectionalized buttons. The numbers of tie-line buttons are identical
to number of tie line switches. The proper combination of tie-line
switches and sectionalized switches without disturbing the radial
nature of distribution system. In case of practical distribution systems, reorganization is much difficult to process with trial and
error method. Hence an alternative approach is required for elevated practical bus systems. Since the improvement algorithm is
employed, so target function has to outline. This work focuses on
reduction of losses for reorganization. The losses of the system further reduced by installing the compensating devices with in eye for
enhancement of savings. This segment is additionally partitioned
into three like load flows, PSO-SSA optimization approach with
their algorithms and flow charts Table 2 Table 3.
PL;T ¼
B
X
ð7Þ
I2k :Rk
ð8Þ
k¼1
WherePT,L is the transmission losses in kW.Rkresistance of the kthline in ohm/km.Vq is voltage of qth busIBis Bthpath current.IL is the
Lthnode current.ZBis Bthpath impedance in ohms/km.
2.1.1. Reorganization network
DNR is a most conventional approach which minimizes the network disturbances. Mainly in distribution structure, the reorganization crisis is to locate most exceptional radial network pattern
that offers smallest amount of power loss at the same time as
employing limitations.
Load stabilizing in to the RDN can be achieved by choosing suitable position of tie line links as the nodes are packed with constant.
Hence, it is imperative to choose a correct location of tie line link.
For practical radial systems, selection of tie line link is very tough
as it has larger nodes. The reasonable of the open buttons could
turn out to be extremely simpler by acquiring this system predicament to the heuristic optimization techniques. The combination of
such improvement procedures used to give ideal outcomes contrasted with natural streamlining strategies, for example, GA,
PSO, and so forth. Hence, this work focuses on combinations of
PSO and SSA algorithms to optimize type-I & II DG reorganization
and most favorable position for utmost economy.
Table 2
IEEE-119 bus RDS Results with and without DG position along with reorganization.
S.
No
1
Tie-line switches
2
3
4
P loss (KW)
% PLoss reduction
Lowest voltage
(p.u)
DG position
(node No)
Type-1 DG Size
(kW)
Type-2 DG size
(kVar)
Loss Cost (USD)
5
6
7
7
Base method
Reconfiguration (devoid of DG)
Reconfiguration (With DG)
132–131-130–129-128–127-126–125-124–123-122–
121-120–119-118
1299.380
25–42-23–50-121–95-58–39-97–7174–129-34–109
881.360
25–42-23–50-121–95-58–39-97–7174–129-34–109
425.280
0.8688
0.9323
0.9323
–
–
116(Type 1) 116(Type-2)
–
–
1138
–
–
1026
6,82,954.128
4,63,242.816
2, 21,244.768
2
Materials Today: Proceedings xxx (xxxx) xxx
V. Rafi and P.K. Dhal
Table 3
IEEE-136 bus RDS Results with and without DG position.
S.
No
Normal method
Reorganization (devoid of DG)
Reorganization (With DG)
156–155-154–153-152–151-150–149-148–147146–145-144–143-142–141-140–139-138–137136
322.440
–
135–128-126–106-118–151-137–71-39–
138-141–58-62–98-145–144-84–148-90–
147
218.711
135–128-126–106-118–151-137–71-39–
138-141–58-62–98-145–144-84–148-90–
147
105.530
Lowest
voltage (p.
u)
DG Position
(node No)
Size of DG
0.9307
0.9616
0.9616
Loss Cost
(USD)
1,69,474.46
1,14,954.50
117
119
1.8291(MW)
1.0824(MVar)
54, 901.68
1
Tie-line
switches
2
3
P loss (KW)
% PLoss
reduction
4
5
6
7
5 Calculate the minimum value from all particles at generations.
6 Update the particles with inertia M.
U ðUpdateÞ ¼ M X þ V i randð1Þ ðPbest U Þ þ V f
randð1Þ ðGbest U Þ
ð10Þ
Where
M ¼ M max ðM max Mmin Þ
ðiter 1Þ
ðng 1Þ
ð11Þ
Fig. 1. A Typical 6-node RDS.
7 Increment the generation until maximum generations.
8 The optimum value is obtained at the end of maximum
generations.
2.2. Particle swarm optimization (PSO)
In a few nonlinear issues, it is a primary swarm-based strategy
as GA has its individual cut-off points on transformation, crossing
over occurrence, time and code unpredictability. GA is to a great
extent subject to the quantity of chromosomes can be expressly
focused to boost the issue. Generally engineering regulations
specifically are of a nonlinear. Phrasings that are utilized in PSO
are inertia, generations, and particles in least and utmost, first
and final velocities. The particles are inhabited with the arbitrary
qualities separate to imperatives and confinements. The particles
are replacements in the target work in every origination. Every
origination the particles are refreshed with least inertia, utmost
inertia, first and last velocities. So, the algorithm for PSO is given
as follows.
2.2.2. Salp swarm algorithm (SSA)
SSA is natures inspired algorithm which is likewise propelled
from swarm-based algorithm. The significant idea driving
swarm-based improvement is looking for the food area. The area
of food location is equivalent to swarm locations at improved spot.
However, the phases of each swarm calculations are non-identical
from their regular behavior of piercing for the food. In SSA likewise,
dragons are called looking through operators, which changed the
areas as per the food locations alongside getting away from predominant foes. Nomenclatures utilized in SSA are looking through
specialists, greatest generations, partition, arrangement, union
food source, adversary position. The algorithm of SSA is as follows:
1. Initialize all the SSA parameters.
2. Populate the searching agents with the variables constraints of
the objective function.
2.2.1. Algorithm
1. Initialize the PSO parameters such as least inertia (Mmin),
utmost inertia (Mmax), first and last velocities (Vi and Vf).
2. Populate the particles within the constraints.
3
u11 u1d
..
.. 7
6 .
U ¼ 4 ..
.
. 5
2
3
s11 s1d
..
.. 7
6 .
S ¼ 4 ..
.
. 5
2
up1
sa1
ð9Þ
ð12Þ
sad
Wheres represents agent searching.s11 represents agents searching
at 1st row and 1st column of the matrix.a denotes numbers of
agents searching.d denotes dimension of the fobj used in SSA.
upd
whereu represents the particle.u11 represents particle at 1st row
and 1st column.uld represents parttcle at 1st row and dth column.
p denotes the number of particles.d denotes dimension of the objective function for solving through PSO.
4 While true.
5 Calculate the Fobj for all agents searching values.
6 Update the position of the agents searching withlow changes
along with other parameters.
7 Determine the optimum fobj value at every generation.
4 At first generation calculate objective function (fobj) at every
particle.
3
V. Rafi and P.K. Dhal
Materials Today: Proceedings xxx (xxxx) xxx
Fig. 2. 119 node RDS along with Tie line switches.
8 Repeat above steps for all generations and fobjopt among all
generations.
9 While end.
3. Proposed technique
The approach which is explained in previous segment is united
and forms the complete proposed philosophy. The reshape and
placement of DGs with combinational optimization is the proposed
methodology for the paper. The reorganization is optimized with
PSO algorithm and DGs are optimally located with SSA. The segment is sub-divided with demonstrating of DGs and hybrid
improvement calculation Fig. 2.
3.1. DG modeling
3.1.1. Type-1 DG modeling
Distributed generators are erected in scattering networks that
are straightforwardly delivering the produced power to load.
Fig. 3. Modeling of DG in radial distribution system.
4
Materials Today: Proceedings xxx (xxxx) xxx
V. Rafi and P.K. Dhal
3.2. Proposed algorithm
Molding of DGs are explained the behavior of the device at the system. Type I DGs are supplying real current to the load which is connected at that bus. Fig. 3 clearly reflects, DG is providing actual
current to load Fig. 4.
The algorithms which are detailed in the prior segment are
combined to explain the optimization problem, for reorganization
of RDS and placement of DG to make best use of the savings in RDS.
Load flow algorithm has been used at each optimizing algorithm
such as PSO and SSA for reorganization and placement of DG.
The PSO terms are adopted to RDS for reorganization. The dimensions of PSO for tie-line switches in radial distribution network.
The particles are populated with dimensions and number of particles. The amount of branches of RDS is always identical to (Nb-
3.1.2. Type-2 DG modeling
The modeling of Type-2 DG is just similar to the Tpye-1 DG. In
Type-2 DG, wattles power of load is compensated by DG. So, functioning of Type-2 DG is that it injects the required wattles power to
the load, which will enhance the voltage outline of the system
Fig. 5.
Fig. 4. Single line graph of 119 bus RDS with reconfiguration.
5
V. Rafi and P.K. Dhal
Materials Today: Proceedings xxx (xxxx) xxx
Fig. 5. Single line structure of IEEE 136 bus RDN.
and type 2 DGs for reducing loss cost along with installation cost
of DGs. So the objective functions are given as follows:
Nmf). Where Nb is no. of nodes and Nmf is no. of main feeders
linked to the grid. Examples of that are IEEE 16 bus system, IEEE
119 bus network and IEEE 136 bus system. IEEE 16 bus system
has three feeders which are connected to the main feeders. So,
numbers of branches are 13 and number of tie-line switches also
three loops are formed with three tie-line switches. IEEE 119 bus
system one main feeder connected to the buses so number of arms
are 118 and number of tie-line switches are fifteen (15) , 136 bus
system one main feeder connected to the buses so amount of
branches are 135 and amount of tie-line switches are twenty one
(21). The reconfigured distribution system is given as input for
the SSA, for best possible situation of DG in radial distribution network. The searching agents are initialized with the two dimensions
which are placement i.e bus number and size of DG. The least and
largest constraints of DG are considered as 10% and 80% of system
load. The limits for bus location are least bus and maximum bus
number of radial distribution system.
The PSO and SSA parameters which are used to formulate the
algorithm is shown in the below table.
The objective function which used for this hybrid algorithm is
applied for two stages. In Istpart, PSO is wield to determine the
optimal reorganization for minimum losses and where as in
2ndpartSSA is wield to determine the optimal position of type-1
f objpso ¼ min Preal kl ð=kWÞ
ð13Þ
0
B
f objDA ¼ minP real kl B
@
Ndg
P
kWhÞþ
i¼1
DGi ðkdg
ð14Þ
=kWhÞÞ
kWÞþkm ð
WherePrealactive power losses in kW.loss cost coefficient kl is 0.06
$/kWh.Installation cost coefficient kdg is 400 $/kW.Maintenance
cost coefficient km is 0.045 $/kWh.
4. Results
The performance of the system is identified through two IEEE
standard 119 and 136 bus structure. Outcomes of system are
explained by means of sufferers and voltage outline system. Voltage report of the structure is determined with reorganization and
devoid of reconfiguration, DGs and misfortunes are likewise contrasted comparatively without and reorganization alongside and
devoid of DG. The segment is additionally spllited with IEEE 119
and IEEE 136 bus system.
6
Materials Today: Proceedings xxx (xxxx) xxx
V. Rafi and P.K. Dhal
at 116th bus with the sizes of 1138 kW and 1026kVar respectively,
with the reduction of losses to 425.28 kW with loss cost of USD 2,
21,244.768. The setting up cost of 400 USD/kW. So lastly, the economy of the structure is 2, 22,305 USD /annum with respect to reorganization loss cost. The outline lowest voltage 0.8688p.u
improved to 0.9323p.u.
4.1. IEEE 119 bus RDS
The single line map of IEEE 119 bus structure is shown in Fig. 2.
The 119 RDN having 132 branches and 118 switches. The 119-node
map has 15 tie-switches are 132–131-130–129-128–127-126–12
5-124–123-122–121-120–119-118 at the time of starting. The
losses of the system with only sectionalized switches and all
opened tie-line switches are 1299.38 kW and the loss cost of the
system with 6,82,954.128 USD, with 0.06 USD/kWh. The losses
after reorganization of 25,42,23,50,121,95,58, 39,97,71,74,129,34,
109 and 130 are 881.36 kw and the loss cost of the system with
reconfiguration 4,63,242.816 USD. The reconfigured distribution
system is given as input for most favorable position of DG using
SSA. The most favorable position of Type-1, Type-2 DG are bath
4.2. Test case II: IEEE 136 bus system
The IEEE 136 bus RDS single-line structure appears in the Fig. 6.
The 136 RDN having 156 branches and 135 switches. The 136 RDN
having 156 branches and 135 switches. The 136-node map has
21tie-switches are 156,154,153,152,151,150,149,148,147,146, 14
5,144,143,142,141,140,139,138,137 and 136 at the time of starting.
Fig. 6. . 136 bus structure RDS with reorganization.
7
V. Rafi and P.K. Dhal
Materials Today: Proceedings xxx (xxxx) xxx
The losses of the system with only sectionalized switches and all
opened tie-line switches are 322.44 kW and the loss cost of the
system with 1,69,474.464 USD, with 0.06 USD/kWh. The losses
after reorganization of 135,128,126,106,118,151,137, 71,39,138,1
41,58,62,98, 145,144,84,148,90,147and 150 are 218.711 kw and
the loss cost of the system with reconfiguration 1,14,954.501
USD. The reconfigured distribution system is given as input for
most favorable position of DG using SSA. The most favorable position of Type-1, Type-2 DG are bath at 116th bus with the sizes of
1138 kW and 1026kVar respectively, with the reduction of losses
to 105.53 kW with loss cost of USD 54, 901.68. The setting up cost
of 400 USD/kW. So lastly, the economy of the structure is 56,538
USD /annum with respect to reorganization loss cost. The outline
lowest voltage 0.9307p.u improved to 0.9616p.u.
References
[1] A. Merlin, Search for a minimal-loss operating spanning tree configuration for
an urban power distribution system, Proc. of 5th PSCC, 1975 1 (1975) 1–18.
[2] D. Das, A fuzzy multiobjective approach for network reconfiguration of
distribution systems, IEEE Trans. Power Delivery 21 (1) (2005) 202–209.
[3] K. Nara, A. Shiose, M. Kitagawa, T. Ishihara, Implementation of genetic
algorithm for distribution systems loss minimum re-configuration, IEEE
Trans. Power Syst. 7 (3) (1992) 1044–1051.
[4] R.S. Rao, S.V.L. Narasimham, M.R. Raju, A.S. Rao, Optimal network
reconfiguration of large-scale distribution system using harmony search
algorithm, IEEE Trans. Power Syst. 26 (3) (2010) 1080–1088.
[5] K. Prakash, M. Sydulu, in: June. Particle Swarm Optimization Based Capacitor
Placement on Radial Distribution Systems, IEEE, 2007, pp. 1–5.
[6] Z.W. Geem, Novel derivative of harmony search algorithm for discrete design
variables, Appl. Math. Comput. 199 (1) (2008) 223–230.
[7] D.K. Molzahn, F. Dörfler, H. Sandberg, S.H. Low, S. Chakrabarti, R. Baldick, J.
Lavaei, A survey of distributed optimization and control algorithms for electric
power systems, IEEE Trans. Smart Grid 8 (6) (2017) 2941–2962.
[8] H.R. Esmaeilian, R. Fadaeinedjad, Energy loss minimization in distribution
systems utilizing an enhanced reconfiguration method integrating distributed
generation, IEEE Syst. J. 9 (4) (2014) 1430–1439.
[9] S. Mirjalili, A. Lewis, The whale optimization algorithm, Adv. Eng. Softw. 95
(2016) 51–67.
[10] R. Tonkoski, L.A. Lopes, T.H. El-Fouly, Coordinated active power curtailment of
grid connected PV inverters for overvoltage prevention, IEEE Trans. Sustain.
Energy 2 (2) (2010) 139–147.
[11] Y. Wang, K.T. Tan, X.Y. Peng, P.L. So, Coordinated control of distributed energystorage systems for voltage regulation in distribution networks, IEEE Trans.
Power Delivery 31 (3) (2015) 1132–1141.
[12] Y. Wang, M.H. Syed, E. Guillo-Sansano, Y. Xu, G.M. Burt, Inverter-based voltage
control of distribution networks: a three-level coordinated method and power
hardware-in-the-loop validation, IEEE Trans. Sustain. Energy (2019).
5. Conclusion
The paper has dealt with arrangement of type-1, type-2 DG to
capitalize on the economy of the framework by including the setting up cost of DG. The paper is utilized PSO for reconfiguration,
which is best swarm based calculation for unraveling a wide range
of frameworks, and SSA is utilized for ideal situation for type I,
type-II DG, which repaid both active and reactive power of the
load. The total calculation is appraisal on two test experiments
which are 119 and 136 IEEE bus structures. The economy of the
frameworks is referenced in USD which is adequately saved per
year.
Further Reading
Declaration of Competing Interest
[1] V. Rafi, P.K. Dhal, Maximization savings in distribution networks with optimal
location of type-I distributed generator along with reconfiguration using PSODA optimization techniques, Mater. Today Proceed. (2020).
The authors declare that they have no known competing financial interests or personal relationships that could have appeared
to influence the work reported in this paper.
8