Optimization of fuzzy controllers as real-time power system stabilizers

Abstract

The optimization and implementation of fuzzy controllers to improve the stability of power systems is presented in the paper. Elements from fuzzy set theory were used in the development of two different types of fuzzy controllers. The first one uses a traditional fuzzy logic controller (FLC) as a power system stabilizer. The input and output membership functions were adjusted according to an evaluation index in order to achieve the optimal performance. The initial FLC design with a complete filled rule table (49 rules) can be reduced using neural-network techniques to improve the dynamic performance. The second type of fuzzy controller is a fuzzy-logic-based power system stabilizer (FLBPSS) which has been simulated to determine its optimum settings using new nonlinear fuzzy membership functions. The performance of the optimized FLBPSS has been compared to the FLC and found to be better. The design was implemented on 5 kVA generators in the Power System Laboratory using an IBM-PC with A/D and D/A converters acting as the real-time controller. Both the simulation and implementation results on one-machine and two-machine infinite bus power systems show that the proposed FLBPSS is quite robust for both small and large perturbations and need not be changed for changes in the systems operating conditions and parameters.