International Journal of Science and Research (IJSR)

ISSN (Online): 2319-7064

Index Copernicus Value (2013): 6.14 | Impact Factor (2014): 5.611

Review of Developments in Methods of

Reconfiguration of Distribution Networks
Prof. A. B. Kansae1, Dr. S. R. Gengaje2
1
Walchand Institute of Technology, Solapur, Maharashtra, India

Abstract: Distribution networks are configured radially. Their configurations may be varied with manual or automatic switching
operations so that all of the loads are supplied and reduce power loss, increase system security, and enhance power quality.
Reconfiguration also relieves the overloading of the network components. The change in network configuration is performed by
opening sectionalizing (normally closed) and closing tie (normally open) switches of the network. These switchings are performed in
such a way that the radiality of the network is maintained and all of the loads are energized. Obviously, the greater the number of
switches is, the greater the possibilities are for reconfiguration and the better the effects are. This paper presents a review of
developments and improvements in distribution network reconfiguration methods which includes multi-agent and genetic algorithm
approach.

Keywords: Distribution networks, reconfiguration, genetic algorithm, multiagent.

1. Introduction each work agent’s solution set, and coordinate them until the
number constraint of switching operations was satisfied. The
Distribution systems consist of groups of interconnected whole architecture of the multi-agent system was presented,
radial circuits. The configuration may be varied via switching and the function modules of the work agent and coordination
operations to transfer loads among the feeders. Two types of agent were also discussed.
switches are used in primary distribution systems. They are
normally closed switches (sectionalizing switches) or Proposed novel systematic algorithm for dynamic
normally open switches (tie switches). Both types are reconfiguration of the distribution network minimizes line
designed for both protection and configuration management. losses. The dynamic reconfiguration was decomposed into a
Distribution network reconfiguration is a process that series of static ones, and a series of solution sets were
consists of changing the status of the network switches in calculated. Then a solution was chosen in each set and all the
order to resupply the non-energized areas after a fault chosen solutions belong to different solution sets composed
occurrence, or optimize given criteria. The reconfiguration the dynamic reconfiguration result. The approach was easily
problem of radial power distribution networks may be stated realized via MAS, and the computer simulations showed the
as follows: given a load profile for a distribution network effectiveness. In the algorithm if the maximum allowed
with a number of tie lines and switching points, find a radial number of switching operations were changed, only the
configuration for the network, which minimizes the network coordination need to be carried out again. And in the
losses, supply the critical loads at acceptable voltage levels. coordination process, along with the reduction of the allowed
number of switching operations, the loss increase quantity
In recent years, considerable research has been conducted for and the time interval that abandoned the optimal solution
loss minimization in the area of network reconfiguration of could be both obtained. If the loss increased remarkably the
distribution systems. Distribution system reconfiguration for allowed number of switching operations should be changed
loss reduction was first proposed by Merlin and Back. They larger than the current value. Conversely, if the loss increased
have used a branch-and-bound-type optimization technique only a little, it should be reduced.
to determine the minimum loss configuration.
The distribution network re-configuration is a process that
The aim of this paper is to present review of developments consists of changing the status of the network switches for re-
and improvements in distribution network reconfiguration routing the power after a fault occurrence, or to optimize
methods which includes multi-agent and genetic algorithm some given criteria. Traditionally, feeder reconfiguration is a
approach. complex combinatorial and constrained optimization problem
because of the numerous combinations of candidate switches.
2. Multi-Agent Technology Chauhan et al. [2] proposed a multi-agent based system for
distribution network reconfiguration. The system consists of
Li et al. [1] proposed a novel systematic algorithm for the load agents and switch agents located at the loads and
dynamic reconfiguration of the distribution network based on switches in the network and a global agent located at the
multi-agent system. In which, one day was divided into substation. The objective was to supply the critical loads
several time intervals and each was managed by a work while maintaining an acceptable system voltage profile.
agent. An improved hybrid particle swarm optimization Simulation results presented for a typical distribution system
algorithm was applied in each time interval to produce a named Circuit of the Future (CoF), developed by Southern
solution set which included some enough good solutions. California Edison (SCE).
Then the coordination agent would select a solution from
Volume 5 Issue 1 January 2016
www.ijsr.net
Paper ID: NOV152676 442
Licensed Under Creative Commons Attribution CC BY
 International Journal of Science and Research (IJSR)
ISSN (Online): 2319-7064
Index Copernicus Value (2013): 6.14 | Impact Factor (2014): 5.611
The approach deploys multi-agent system in conjunction with this approach is demonstrated on the example of a real cable
graph theory algorithms to reconfigure the system. The MV distribution network.
proposed Multi-agent system makes use of both centralized
and decentralized methods to make up for the disadvantages The Figure 1 shows the trends of penalizations and SAIDI
of each method. The proposed model of MAS is tested on values during optimization. Average values of SAIDI for
proto-type distribution network named Circuit of the Future Portuguese (PT) and Modified Portuguese (X) are very
(CoF). The team is developing multi-agent deployments and similar, on the contrary to Finish (FI) penalization model.
application algorithms to reconfigure the system to supply the Finish model also has the highest value of penalization and
critical loads and to maintain the good voltage profile across modified Portuguese achieves low order values compare to
the network even in the event of faults. The fault others standards. Considering the results, graphs reveal that
reconfiguration algorithm makes use of shunt compensation SAIFI and SAIDI values do not have to correspond directly
and priority based load shedding in order to control the to the values of penalizations. The solution with minimal
voltage. The fault reconfiguration algorithm proposed in this penalization does not determine that SAIDI and SAIFI are
paper chooses the best possible switching configuration in minimal as well. Moreover the results showed that different
regards to voltage profile and real power loss in the system. penalization models lead to various optimal outputs. These
The results obtained are very promising and show superior conclusions are in accordance with results.
ability of Multi-agent systems in the field of fault detection
and reconfiguration. The next stage of the work will be
focused on reconfiguration in a even more complicated smart
system and finding better interface between different
simulation platforms for efficiency improvement.

Lin et al. [3] proposed a novel method is proposed to solve

the distribution network reconfiguration. The proposed
method combines the binary particle swarm optimization
(BPSO) with discrete particle swarm optimization (DPSO)
and multi-agent system (MAS). It also incorporates self-
learning mechanism and the survival of the fittest which is a
part of natural selection for reconfigure distribution network.
In the algorithm, each particle is thought to be an agent and
all agents live in a lattice-like environment. The competition
and cooperation operator is preformed on every agent. If the
agent is a loser, a new agent will occupy its lattice-point
which is determined by BPSO and DPSO. If the agent is a Figure 1: Comparison of Penalization and SAIDI among all
winner, it can perform the self-learning mechanism, so as to standards
obtain the better fitness value. Finally the next iteration
particles are determined by the fitness value of previous Sawa et al. [5] presented a new Particle Swarm Optimization
particles, according to the survival of the fittest mechanism. (PSO) for optimal radial network reconfiguration. Network
The result showed that the method has superior features, reconfiguration is a combinatorial problem. The combination
including good computation efficiency, good convergence of search space increases exponentially with the number of
characteristics, and high-quality solutions etc. switches. Fixed loop coding structure is introduced to reduce
the number of combinations. The feature of this method is to
In MAS, to build a lattice-like environment, all agents introduce mutation operation for discrete decimal problem to
improve their communication by the competitive and PSO. This method, discrete decimal PSO with mutation
cooperative operation, unlike individual agents in PSO. In (DDM-PSO), is applied to a test system with 37 nodes and
this process, it is in favor of converged to the global 63 branches. The proposed method improves the maximum
optimized results. Those particles will divide into two parts arrival number at the optimal solution from 8 to 17 compared
through the competition and cooperation operator. The losers to PSO without mutation in twenty trials. This number of
will update by HPSO, which can avoid premature DDM-PSO is also more than that of GA, 10.
convergence and to get the best solution in a global search.
The winners will update by self-learning mechanism. An innovative radial network reconfiguration method has
been developed using the discrete decimal mutant PSO
3. Genetic Algorithm (DDM-PSO). The feature of this method is to introduce
mutation operation and a fixed loop coding structure. This
Paar et al. [4] presented one possible solution to find optimal method was applied to a test system with 37 nodes and
configuration of switches in the meshed distribution network 63branches. The proposed method improves the maximum
which leads to minimal interruption payment for selected arrival number at the optimal solution from 8 to 17 compared
continuity standard - Finnish, Portuguese and modified to PSO without mutation. This showed the effectiveness of
Portuguese. For its suitable properties, simple genetic the developed method.
algorithm was chosen as the optimization. Functionality of
Reconfiguration of radial distribution system is the
significant way of altering the flow of power through lines.
Volume 5 Issue 1 January 2016
www.ijsr.net
Paper ID: NOV152676 443
Licensed Under Creative Commons Attribution CC BY
 International Journal of Science and Research (IJSR)
ISSN (Online): 2319-7064
Index Copernicus Value (2013): 6.14 | Impact Factor (2014): 5.611
This altered flow changes the real power losses, reactive Agent System”, IEEE Xplore, page no. 1305-1311,
power losses and voltage profiles. Privatized RDS need to 2008.
operate profitably with minimum operational losses and [2] Sridhar Chouhan, Hui. Wan, H.J.Lai, Ali Feliachi, M. A.
power quality. Envisaging such a prospect, Venkatesh et al. Choudhry, “Intelligent Reconfiguration of Smart
[6] focused on the aspects of loss minimization and voltage Distribution Network using Multi-Agent Technology”,
enhancement of RDS by artificial intelligence methods. A IEEE Xplore, 2009.
sample 33-bus system and 69-bus system are chosen for the [3] Hongbin Sun, Yongsheng Ding, “Network
study and the results compared. Reconfiguration of Distribution System Using Fuzzy
Preferences Multi- Objective Approach”, 2nd
This method used two algorithms namely FMGA and EP to International Asia Conference on Informatics in Control,
reconfigure the RDS by minimizing the real and reactive Automation and Robotics, page no. 93-96, 2010.
power losses and at the same time improving the power [4] Martin PAAR, Pekka VERHO, Petr TOMAN, “An
quality. The comparison of results with both the algorithms Application Of Genetic Algorithms To The Distribution
showed that FMGA is able to find potentially good switching Network Reconfiguration For Minimization Of
options. EP has good solution but it takes a longer time to Interruption Costs”, 20th International Conference on
converge. In FMGA, more potential solutions that are found Electricity Distribution, 2009.
quickly and crossover exploits these new candidate solutions [5] T. Sawa, “Radial Network Reconfiguration Method in
to find the best solution quickly. Distribution System using Mutation Particle Swarm
Optimization”, IEEE Bucharest Power Tech Conference,
The topology of an Electrical Distribution System (EDS) can page no. 1-6, 2009.
be suitably modified to minimize the real losses. Basically, [6] B. Venkatesh, S. Chandramohan, N. Kayalvizhi, R.P.
losses in an EDS arise due to two factors: 1) fault in the Kumudini Devi, “Optimal Reconfiguration of Radial
network 2) overload in the feeders. EDS is normally Distribution System Using Artificial Intelligence
unevenly loaded and hence often need load balancing, which Methods” page no. 660-665, 2009.
can also be done by reconfiguring the network like changing [7] P. Ravibabu, M.V.S.Ramya, R.Sandeep, M.V.Karthik,
the status of both sectionalizing and Tie switches. They S.Harsha, “Implementation of Improved Genetic
presented a new approach for optimal reconfiguration of a Algorithm in Distribution System with Feeder
radial Electrical Distribution Network based on the advanced Reconfiguration to Minimize Real Power Losses”, 2nd
genetic algorithm using improved methods of selection, International Conference on Computer Engineering and
crossover and fitness function to determine the optimal Technology, Vol. 4, page no. 320-323,2010.
configuration path. The opened switches are taken into
consideration for crossover process. After obtaining number Author Profile
of solutions (off springs) from the combinational analysis, the
optimal solution is selected based on the fitness function, i.e., Prof. A. B. Kanase received his M.E in Electronics
the solution having the minimal power losses. Engineering from Shivaji University, Kolhapur,
Maharashtra State, India. He is currently working as
Ravibabu et al. [7] presented an advanced method to study Principal in Karmayogi Polytechnic College,
the electrical distribution network through feeder Pandharpur, Dist. Solapur, Maharashtra State, India.
reconfiguration using Improved Genetic Algorithm (IGA). His research interest includes Power Electronics, Optimization
The Distribution Network Feeder Reconfiguration (DNRC) is Techniques and Automation.
modified efficiently using some improvements in crossover
Dr. S. R. Gengaje received his PhD in Electronics
and mutation. From several case studies and comparison with
Engineering from Shivaji University, Kolhapur,
other methods including heuristic approach, it is proved that Maharashtra State, India. He is currently working as
Improved Genetic Algorithm is more efficient and satisfies Professor and Head, Department of Electronics in
the conditions of the global optima. Walchand Institute of Technology, Solapur,
Maharashtra State, India. His research interest includes Soft
4. Conclusion Computing, Image and Video Processing and Telecommunication.

This paper presents review of developments in

reconfiguration methods of distribution networks. The results
presented showed that the multiagent systems have superior
ability in the field of fault detection and reconfiguration. A
study on electrical distribution using improved Genetic
Algorithm is proved to be efficient and satisfies the
conditions of the global optima.

References
[1] Zhenkun Li, Xingying Chen, Kun Yu, Bo Zhao, and
Haoming Liu, “Novel Approach for Dynamic
Reconfiguration of the Distribution Network via Multi-

Volume 5 Issue 1 January 2016

www.ijsr.net
Paper ID: NOV152676 444
Licensed Under Creative Commons Attribution CC BY