Neuro-fuzzy based power system stabilizer of a multi-machine system

Abstract

This thesis presents a study of neuro-fuzzy power system stabilizer (PSS) for stability enhancement of a multi-machine power system. In order to accomplish a stability enhancement, speed deviation (??) and acceleration (? ) of the rotor synchronous generator were taken as the input to the neuro-fuzzy controller. These variables take significant effects on damping the generator shaft mechanical oscillations. The stabilizing signals were computed using the neuro-fuzzy membership function depending on these variables. Simulink Block Design and Matlab 7.6 were utilized in implementing the study. The simulations were tested under different types of conditions; the steady state operation, the three phases to ground fault, with load connected to the system, mechanical input power changes and reference voltage (Vref) step changes. The performance of the neuro-fuzzy power system stabilizer was compared with the conventional power system stabilizer and without power system stabilizer