FUZZY LOGIC BASED CONTROLLERS FOR POWER SYSTEMS

Abstract

Abstract : In this thesis, the application of fuzzy logic control to two control problems in electrical power systems has been investigated. To demonstrate the effectiveness of fuzzy-based controllers, a comparison is made with other controller designs based on classical and modern control theories. The first control problem that has been investigated is low-frequency oscillation. Various power system stabilizers (PSSs), based on different control methodologies, have been studied for a single-machine infinite-bus system to dampen oscillations and improve system dynamic performance. These PSSs are, namely, the conventional lead stabilizer based on classical control theory, a linear optimal state regulator, a linear suboptimal output regulator, rule-based stabilizers, and various types of fuzzy logic stabilizers. All of these stabilizers have been designed under the same operating conditions, and their performance has been evaluated at nominal and off-nominal operating conditions when the system is subjected to small disturbances in terminal voltage or mechanical torque. In addition, the effect of variations in some system parameters on the performance of the stabilizers has been investigated. The digital computer simulation results show the effectiveness of the fuzzy logic-based stabilizers over the others The second control problem studied is the load frequency control of interconnected systems. Different load frequency controllers (LFCs) have been designed for the two-area power system to maintain the frequency and power exchange over inter-area tie-lines at scheduled values. The load frequency controllers investigated are the conventional integral LFC, the linear optimal state regulator LFC, and four other newly proposed fuzzy-based LFCs. The first proposed fuzzy controller was constructed using 49 rules with area control error (ACE) and its rate as input signals, while the second proposed fuzzy controller was constructed using only 9 rules with area frequency deviation and tie-line power deviation as input signals. The other two newly proposed LFCs are a conventional integral controller with fuzzy gain and optimal scheduling state regulation with fuzzy gain scheduling. The fuzzy scheduling is done using sugeno-type of fuzzy rules each of which represents a particular operating condition. These two proposed controllers are adaptive with respect to the system operating condition characterized by power system time constant, tie-line synchronizing coefficient and the frequency bias setting. All of these LFC's are compared in terms of their performance and effectiveness in maintaining the two area frequency as well as the tie-line power exchange at their scheduled values ​​following a step load disturbance in area demand at nominal and off-nominal operating system conditions. Simulation results show that the proposed LFC with fuzzy gain scheduling yield better dynamic performance than the other LFC's particularly at off-nominal system operating conditions.