IMPLEMENTATION OF PROBABILISTIC NEURAL NETWORK (PNN) BASED AUTOMATIC VOLTAGE REGULATOR (AVR) FOR EXCITATION CONTROL SYSTEM IN MATLAB

In an interconnected systems it is quite difficult to maintain terminal voltages and frequency in specified limits. The function of load frequency controller (LFC) and AVR is to maintain frequency and terminal voltages constant so as to achieve synchronism. The high gain and fast action of traditional AVRs introduce negative damping which causes oscillations in power system, to overcome this sophisticated issue power system needs self-tuned AVRs. This research work focuses on design and implementation of PNN based AVR to enhance the transient stability limit of power system. In this work, AVR is designed using different control schemes is in MATLAB/ Simulink. Performance of AVR is analyzed using conventional controller, PID controller and PNN controller. Simulation results are presented using time domain simulations. Results witnesses the robustness of PNN based AVRs over a PID and conventional AVRs in terms of reduced overshoot value, settling time and rise time.