ABSTRACT
This worked aimed at developing an optimal reconfiguration algorithm for a radial distribution
network using fast and elitist non-dominated sorting genetic algorithm (NSGA II), considering
distributed generation. The work models the reconfiguration using a pragmatic multi-objective
approach considering active power loss and total voltage deviation, so as to determine the optimum
locations of the tie and sectionalizing switches within the distribution network. The model was
validated using standard IEEE 33-Bus network and extended to a subsection of Zaria distribution
network. The active power loss and total voltage deviation was estimated for Gaskiya, Railway,
Sabo and Canteen distribution network as 55.32kW, 0.22V, 17.22kW, 0.2340V, 120.08kW,
0.9949V and 508.0kW, 4.7482V respectively, prior to reconfiguration. With distributed generation
placed at different location based on the computed voltage stability index of the nodes. The active
power loss for Gaskiya, Railway, Sabo and Canteen distribution network was recorded to be
26.19%, 19.22%, 10.23% and 8.01% reduction respectively as compared to the initial
configuration with distributed generation placed at strategic locations. The optimal location of the
tie and sectionalizing switches for Gaskiya, Railway, Sabo and Canteen distribution network was
found to be 12 14, 14 16 18, 25 20 11 and 23 16 43, respectively after reconfiguration. A reduction
in active power loss and reductions in total voltage deviation for Gaskiya, Railway, Sabo and
Canteen distribution network was found to be 37.64%, 28.94%, 18.2%, 8.12%, 39.21%,37.8% and
23.42%, 10.72% respectively as compared to the active power loss and total voltage deviation of
the initial configuration.
1
TABLE OF CONTENTS
DECLARATION ……………………………………………………………………………………………………………. ii
CERTIFICATION …………………………………………………………………………………………………………. iii
DEDICATION ………………………………………………………………………………………………………………. iv
ACKNOWLEDGEMENTS ……………………………………………………………………………………………… v
TABLE OF CONTENT ………………………………………………………………………….vi
LIST OF TABLE………………………………………………………………………………… ix
LIST OF FIGURE ……………………………………………………………………………… .x
LIST OF ABBREVIATION……………………………………………………………………..xi
ABSTRACT ……………………………………………………………………………………………………………….. xiiii
CHAPTER ONE: INTRODUCTION ………………………………………………………………………………… 1
1.1 Background …………………………………………………………………………………………………………… 1
1.2 Statement of the Problem ………………………………………………………………………………………… 5
1.3 Aim and Objectives ………………………………………………………………………………………………… 5
1.4 Methodology …………………………………………………………………………………………………………. 6
1.5 Significance Contribution ……………………………………………………………………………………….. 7
1.6 Thesis Organization………………………………………………………………………………………………… 7
CHAPTER TWO: LITERATURE REVIEW ……………………………………………………………………… 8
2.1 Introduction ………………………………………………………………………………………………………………. 8
2.2 Conventional Approach ……………………………………………………………………………………………… 8
2.2.1 Determination of Radial Configuration…………………………………………………………………… 8
2.1.2 Kirchhoff Matrix Tree Theorem ……………………………………………………………………………. 9
2.1.3 Heuristic Approach to Distribution Network Reconfiguration …………………………………. 10
2.1.3.1 Branch-and-Bound Technique ……………………………………………………………………….. 10
2.1.3.2 Optimal Flow Pattern ……………………………………………………………………………………. 11
2.1.3.3 Ruled-Based Comprehensive Approach ………………………………………………………….. 12
2.1.4 Graph Theory Concept and Application to Reconfiguration ……………………………………. 12
2.1.5 Mixed Integer Linear Programming Approach ………………………………………………………. 13
2.1.5.1 Simplified Mathematical Model of Distribution Network Reconfiguration …………. 13
2.1.5.2 Mixed-Integer Linear Model …………………………………………………………………………. 16
viii
2.1.6 Application of Genetic Algorithm to Network Reconfiguration ………………………………. 17
2.1.7 Spanning Tree …………………………………………………………………………………………………… 20
2.1.8 Matroid Theory to Reconfiguration ……………………………………………………………………… 20
2.1.9 Depth First Search Algorithm ……………………………………………………………………………… 22
2.1.10 Overview of Meta-Heuristic Technique for Network Reconfiguration ……………………. 22
2.1.10.1 Application Ant Colony optimization for Network Reconfiguration …………………. 22
2.1.10.2 Application of Particle Swarm Optimization for Network Reconfiguration ……….. 23
2.1.11 Application of Multi Objective Evolutional Programming for Network Reconfiguration
………………………………………………………………………………………………………………………………… 25
2.1.11.1 Sum of Weighted Cost Function ………………………………………………………………….. 25
2.1.11.2 Multi objective Evolution Programming Using Fuzzy Objective Function ………… 26
2.1.11.3 Pareto Optimal Concept ………………………………………………………………………………. 28
2.1.11.4 Fast and Elitist Non-Dominated Sorting Genetic Algorithm (NSGA II) ……………. 28
2.2 REVIEW OF SIMILAR WORKS ……………………………………………………………………………… 31
CHAPTER THREE:MATERIALS AND METHODS ………………………………………………………. 37
3.1 Introduction …………………………………………………………………………………………………………….. 37
3.2 Model Equation of the Two Bus Distribution Network…………………………………………………. 37
3.3 Load Model …………………………………………………………………………………………………………….. 40
3.3.1 Constant Power Load (CP) ………………………………………………………………………………….. 41
3.3.2 Constant Current Load (CI)…………………………………………………………………………………. 41
3.3.3 Constant Impedance Load (CZ) …………………………………………………………………………… 41
3.3.4 Exponential Load ………………………………………………………………………………………………. 41
3.4 Modelling of Distributed Generation ………………………………………………………………………….. 42
3.4.1 Constant Active and Reactive Power Model………………………………………………………….. 42
3.4.2 Constant Active Power and Constant Voltage Model ……………………………………………… 43
3.5 Distributed Generation placement ……………………………………………………………………………… 44
3.6 Developed Approach ……………………………………………………………………………………………….. 44
3.6.1 Data Collection ………………………………………………………………………………………………….. 44
3.6.2 Determination of Initial Configuration …………………………………………………………………. 45
3.6.2 Formulation of the Problem ………………………………………………………………………………… 47
3.7 Conclusion ……………………………………………………………………………………………………………… 51
ix
CHAPTER FOUR:RESULTS AND DISCUSSIONS ………………………………………………………… 52
4.1 Introduction …………………………………………………………………………………………………………….. 52
4.2 Comparative Study Before and After the Application of Reconfiguration ………………………. 52
4.2.1 Gaskiya 16-Bus Network ……………………………………………………………………………………. 54
4.2.2 Railway 19-Bus Network ……………………………………………………………………………………. 55
4.2.3 Sabo 29-Bus Distribution Network ………………………………………………………………………. 55
4.2.4 Canteen 50-Bus Network ……………………………………………………………………………………. 56
4.2.5 Discussion ……………………………………………………………………………….55
4.3 Analysis of the Voltage Profile ………………………………………………………………………………….. 60
4.3.1 Gaskiya 16-Bus Network ……………………………………………………………………………………. 60
4.3.2 Railway 19-Bus Network ……………………………………………………………………………………. 61
4.3.3 Sabo 29-Bus Network ………………………………………………………………………………………… 63
4.3.4 Canteen 50-Bus Network ……………………………………………………………………………………. 64
4.4 Conclusion ……………………………………………………………………………………………………………… 66
CHAPTER FIVE: CONCLUSION AND RECOMMENDATIONS ……………………………………. 67
5.1 Introduction ………………………………………………………………………………….63
5.2 Conclusion ……………………………………………………………………………………………………………… 67
5.3Limitations ………………………………………………………………………………………………………………. 68
5.4 Recommendation for Further Work ……………………………………………………………………………. 68
Reference …………………………………………………………………………………………………………………….. 69
Appendix …………………………………………………………………………………………68
x
CHAPTER ONE
INTRODUCTION
1.1 Background
The concept of distribution system automation has capture the attention of researchers and utility
companies over the last decade. Even though, a lot of research have been undertaken and many
works are still ongoing at the moment on “distribution system automation”, due to technological
innovations and increasing connection of distributed energy source in to existing distribution
network. The research is aimed at developing an optimal reconfiguration model for a radial
distribution network using Zaria distribution network as a case study.
The problem of minimizing distribution systems losses has received a global attention due to high
cost of electrical energy, need for better quality of service, efficient utilization of available energy
and high power loss in power networks (Charlangsut et al, 2012). Under normal operating
conditions, any distribution network must supply electrical power to all its customer connected to
it, while simultaneously avoiding overloading, feeder thermal overload and abnormal voltage
across the line, as well as minimizing active power loss and maintaining radial topology (Carcamo-
Gallardo et al, 2008) etc. There are different techniques for reducing losses within the distribution
level which include reconfiguration, capacitor placement, load balancing, introduction of higher
voltage level and reconductoring (Abdelaziz et al, 2010; Sarfi et al, 1994). These methods of
reducing losses are quite numerous but the major concern is about their technical implications on
the network. Introduction of new equipment at distribution level offers tremendous financial
burden on the utilities that cannot be justified by the potential saving. The use of fixed
2
compensators offers optimal reduction in losses for specific demand condition, but the control
systems required are very expensive (Sarfi et al, 1994).
The choice of reconductoring is not an option due to the cost associated with relaying the feeder,
while the introduction of higher voltage level requires upgrading the rating of the transformer and
other equipment in the network (Sarfi et al, 1994). The problem of reconfiguration of distribution
network has been explored in this work due to its flexibility, ability to use existing equipment and
other pragmatic approaches to offer dynamic means of reducing losses within the distribution
level.
Distribution systems are designed as radial systems in which tie (NO) and sectionalizing
switches(NC) play a critical role in determining the configuration of the system (Rashtchi and
Pashai, 2012). Therefore reconfiguration of the distribution network is the process of altering the
topology of the distribution network implemented via maneuvering the position of both switches
(normally closed and normally open). Due to the candidate switching combinations in the
distribution system, reconfiguration is a complicated combinatorial, non-differentiable constrained
optimization problem (Abdelaziz et al, 2010).
The reconfiguration problem is one of the multi criteria and multi objective optimization types,
where the solution is chosen after the evaluation of some indices such as active power loss,
reliability indices, branch load limit and voltage drop limit which represent multiple functions.
These criteria can be grouped into objective functions that can be minimized and the constraints
that must be included with some bounds while some criteria are often modeled along with their
objective functions (Tomoigo et al, 2013). The earlier reconfiguration problems were mostly
represented by a single objective function (Abdelaziz et al, 2010; Baran and Wu 1989; Nara et al,
3
1992). The growing demand and increasing connection of smart devices in to the existing network
offers a great advantage but increases the burden and complexity of distribution network.
Clearly, single-shot optimization is not an effective solution to the problem due to the changes in
power demand of the customer over time. This is because an operating topology obtained at a
specific operating condition may no longer be optimal under new operating conditions (Carcamo-
Gallardo et al, 2008). There is the need to constantly evolve efficient techniques of reducing active
power losses within the distribution system in order to enhance the network performance while
improving the continuity of electricity delivery to consumer. The early works on reconfiguration
of radial distribution network were centralized in planning of distribution system with the aim of
reducing system cost. Distribution network planning requires feeder reconfiguration on a seasonal
and annual basis, with the potential to provide required benefit, since the load varies continuously
as the network expands. This potential can be exploited by distribution management system where
the switches can be opened and closed on a real-time basis (Roytelman et al, 1996). The concept
of reconfiguration of radial distribution network is aimed at improving the system performance
under different operating conditions.
4
Figure 1.1 Diagram of Typical Radial Distribution Network (Zhang et al, 2012)
Generally, distribution networks are built as interconnected networks, while in operation they are
arranged in a radial tree structure. This means that a distribution system is divided into subsystems
of radial feeders, which contain a number of normally closed (NC) switches and normally open
(NO) switches as in Figure 1.1. According to the graph theory (Zhu, 2009), a distribution network
can be represented with a graph of G (N, B) that contains a set of N nodes and a set of B branches.
Every node represents either a source node (supply transformer) or a sink node (customer load
point), while a branch represents a feeder section that can either be energized or de-energized. The
typical radial distribution network of Figure 1.1 is such that the feeder section forms a set of trees
where each sink node is supplied from exactly one source node. The distribution network
reconfiguration problem then is to find a radial operating structure that minimizes the system
power loss while satisfying critical operating constraints.
Many techniques such as branch and bound, analytic approach, heuristic, expert system, linear
programming etc were presented by different researchers aimed at addressing the problem of
5
reconfiguration of radial distribution network considering active power loss (Zhu, 2009). Fewer
addressed the problem of optimal reconfiguration using multi-objective approach (Hung, 2002;
Prasad et al., 2009; Kumar et al., 2006; Tomoiga, 2013). This research work developed a pragmatic
multi objective genetic algorithm approach in addressing the problem of reconfiguration of
distribution system in order to minimize active power loss and reduce total voltage deviation.
1.2 Statement of the Problem
Most of the literatures reviewed addressed the problem of reconfiguration using single objective
function considering active power loss as the optimization parameter. Fewer researchers addressed
the problem using multi-objective optimization approach by transforming the multi-objective
function into a single objective function by using some user parameters which are subjective. The
advent and increasing connection of distributed energy sources to distribution network, has
necessitated the need to consider other parameters (such as total voltage deviation, reliability
index, load balancing, energy not served etc) so as to increase the complexity of the system. This
research work addressed the problem of distribution network reconfiguration using an enhanced
non-dominated sorting genetic algorithm (NSGA-II) considering active power loss and total
voltage deviation as the objective functions, within a reasonable convergence time.
1.3 Aim and Objectives
The aim of this research is to develop an optimal reconfiguration algorithm for radial distribution
networks, using non-dominated sorting genetic algorithm (NSGA-II) with a view to improving the
performances of a radial distribution system through minimizing its active power loss and reduce
total voltage deviation.
The objectives of this research are itemized as follows:
6
i. Development of an optimal reconfiguration algorithm using fast and elitist non-dominated
sorting genetic algorithm (NSGA-II) to determine the optimal location of tie and
sectionalizing switches.
ii. Determination of the location and size distributed generation for the distribution network.
iii. Investigation of the effect of distributed generation placement on the power quality for the
distribution network.
iv. Comparison of the results obtained from the normal and reconfigured network in order to
determine the extent of reductions in both active power loss and total voltage deviation
v. Comparison of the efficiency of the proposed model with existing model as a means of
validation.
1.4 Methodology
The following methodology are adopted in carrying out this research and are itemized as follows;
i. Collection of data for radial distribution network of Zaria distribution
ii. Establishment of the most suitable method for performing power flow analysis for the
radial distribution network
iii. Based on ii, a power flow analysis is carried out to estimate the initial configuration for
distribution network.
iv. Determination of the optimal location and size of distributed generation placement based
on computed voltage stability index.
v. Development of an enhanced non-dominated sorting genetic algorithm (NSGA-II) model
used to determine the optimal location of tie and sectionalizing switches considering active
power loss and total voltage deviation.
vi. Testing and validation of the developed model
7
vii. A comparison of the reconfigure system with the original system is carried out, to
determine the extent of loss reduction.
1.5 Significant Contribution
The significance of this research are itemized as follows:
i. Development of a multi-objective based reconfiguration model using enhanced Nondominated
Sorting Genetic Algorithm (NSGA-II) to obtain optimal switching state (open
branch).
ii. A reduction in active power loss (2.8%) was recorded when tested on a standard 33-bus
IEEE network as compared with active power loss of other developed model.
1.6 Thesis Organization
The general introduction has been presented in Chapter One. The rest of the chapter are as follows:
A detailed of the fundamental concepts of reconfiguration and review of the relevant literatures is
carried out in Chapter Two. Mathematical derivation of reconfiguration equations and formulation
of the problem are presented in Chapter Three. Analysis and discussions of the results are presented
in Chapter Four. Conclusion and recommendation are presented in Chapter Five.
8
IF YOU CAN'T FIND YOUR TOPIC, CLICK HERE TO HIRE A WRITER»
