An improvement of frequency performance in a power system optimization via controller parameter design using genetic algorithm

Abstract

An application of proportional-integral (PI) controller and linear quadratic regulator (LQR) optimal control to suppress the settling time of frequency response as the output of the power system in oscillation is proposed as optimal PI controller. In this application, the weighting matrices, the bias factors, the proportional gains, and the integral gains as important controller parameters greatly determine the frequency response of the system. Commonly these controller parameters are designed in constant values, but actually using iterative calculations as modern adaptive plant, can be developed. Therefore, this paper introduces an application of genetic algorithm (GA) as a modern tool to design controller parameters that consist of weighting matrices and bias factors to obtain the proportional and integral gain as an optimal gain of power system in its optimization. Also we propose the applications of integral controller and conventional PI-controller as alternative application. Using the optimal PI-controller, its parameters designed by using constant values and the inverse-square method (ISM) are also demonstrated. The final results show that application of optimal PI-controller designed by using GA on the power system can suppress frequency oscillation of the system in the shortest settling time among the applications of the alternative of proposed controls.