Multilayer Perceptrons Neural Network Automatic Voltage Regulator With Applicability And Improvement In Power System Transient Stability

In electrical power system network, the excitation system contributes in an effective voltage control and enhancement of the system stability. It must be able to respond quickly to a disturbance enhancing the transient stability and the small signal stability. This work presents an artificial neural network (ANN) based automatic voltage regulator (AVR) controller for the excitation voltage system of synchronous machine in order to investigate the applicability and to improve the transient response. Multilayer Perceptrons (MLP) is the most popular type of Feedforward neural networks (FFNN) architecture of ANN and has proved many suceesful applications in power system and power system control and stability. The linearized model of SM with single machine connected to infinite bus (SMIB) and AVR xcitation system is developped in Matlab/Simulink. The performance of proposed MLP neural networks is tested, verified and compared with conventional PID Proportional Integral Derivative (PID) AVR controller of synchronous generator. From simulation results, it is found that the MLP ANN AVR controller demonstrates not only the promising applicability but also better response by removing oscillations very quickly improving transient stability of power system.