Power Stability and Control with Neural Network-Based Modelling

This paper deals with the controller design of synchronous generators founded on artificial neural networks (ANN). It is an investigation into the application of a neuro-controller to the subject of a single-machine infinite bus (SMIB) power system. The SMIB pattern is a fundamental representation of a power system, permitting the evaluation of control strategies to enhance stability. To increase the stability of a single-machine infinite bus (SMIB) power system, this work aims to compare a neural controller with a Power System Stabilizer (PSS) and a proportional-integral-derivative (PID) controller. The neural controller, PSS controller, and PID controller are all utilized to control the generator's excitation system. Their effectiveness is tested by simulating them under many different operating conditions and disturbances. An artificial neural network (ANN) is employed to establish the correlation between the excitation level and the terminal voltage of a generator. Simulations are conducted using MATLAB/Simulink, with the neuro-controller replacing Power System Stabilizer (PSS) or Automatic Voltage Regulator (AVR) controllers. The Levenberg-Marquardt algorithm is used to determine the ANN's optimal weight coefficients. This relationship is established by observing the generator's response to changes in reference voltage levels. The controllers elucidate the actual goodness of the suggestion controller in a period of voltage regulation and transient stability of the electrical power system.