Please use this identifier to cite or link to this item:
Title: Genetic algorithm for optimal capacitor selection and optimal power flow with facts devices
Authors: Leung, Ho-chung
Degree: Ph.D.
Issue Date: 2004
Abstract: As electric utility industry shifts from regulated monopolies to competitive market, it will place an increased reliance on the existing transmission systems. Flexible AC Transmission Systems (FACTS) become more important since they can enhance system flexibility and increase loadability. To address this need, the aim of this research is to develop the use of genetic algorithm for capacitor selection and to develop control strategies for FACTS devices in power systems. The main contributions of this research work are thus to develop computer algorithms using Genetic Algorithms (GA) to solve the optimal capacitor allocation problem with harmonic distortion considerations and the optimal control setting problem of FACTS devices in optimal power flow (OPF). The first part of this thesis reports the research findings of a genetic algorithm approach for optimizing shunt capacitor sizes and their placement in radial distribution systems with the consideration of harmonic distortion limit due to the presence of nonlinear power electronic devices. The algorithm is based on a genetic algorithm (GA) solution technique to minimize cost under the additional constraints of maximum limit in Harmonic Distortion Factor (HDF) and voltage. A harmonic distortion calculation is embedded in the genetic algorithm solution routine to enhance the optimal capacitor allocation solution. Results of simulation show that the approach is effective for such discrete value optimization problem. The improvement of the harmonic distortion is effective and the best allocation of capacitors is selected. Secondly, the thesis would present the development of the equivalent modeling by Power Injected Method (PIM) of various types of FACTS devices including Thyristor Controlled Series Compensator (TCSC), Thyristor Controlled Phase Shifter (TCPS) and Unified Power Flow Controller (UPFC). A real-coded genetic algorithm method is presented to solve the optimal power flow problem of power system with flexible AC transmission systems (FACTS). The proposed method introduces the injected power model of FACTS devices into Newton-Raphson (NR) power flow problem to exploit the characteristic of FACTS devices. The advantage of this method is that it is easily incorporated into existing OPF of Energy Management Systems (EMS) since it would reduce maintenance cost and software development. Moreover, the admittance matrix can be kept constant during the load flow calculation. Case studies on IEEE test systems demonstrate the potential for application of GA to determine the control parameter of the power flow controls with FACTS. It is shown that the FACTS device would not provide significant cost saving since cost depends mainly on the active power flow. However, it can increase the controllability and flexibility of the system; it can provide wider operating margin and improved voltage stability with higher reserve capacity. As deregulation and contract path are becoming more important, FACTS devices play an increasingly important role in such power system operation. It is shown in the thesis that TCSC and TCPS can be employed to control the active power flow while UPFC can be used to control the real and reactive power. They can redistribute the power flow to the available transmission lines within the transmission capacity to achieve more efficient utilization of their capacity. As cost is the main objective of OPF, more constraints will increase the generation cost. In the developed methodology, GA effectively finds the optimal setting of the control parameters using the conventional OPF method as an embedded calculation tool. Overall, the results show that GA is suitable in dealing with non-continuous, non-differentiable and non-convex optimization problems, such as the capacitor selection and optimal power flow problems with FACTS devices.
Subjects: Hong Kong Polytechnic University -- Dissertations
Genetic algorithms
Electric power systems
Pages: xiii, 130 leaves : ill. ; 30 cm
Appears in Collections:Thesis

Show full item record

Page views

Last Week
Last month
Citations as of Jun 4, 2023

Google ScholarTM


Items in DSpace are protected by copyright, with all rights reserved, unless otherwise indicated.