Design of genetic algorithm based optimal fuzzy logic controller for a non linear process

Many practical processes exhibit nonlinear behavior. However, most controller design approaches are based on linear models. Although linear model based control methods are well established, they don´t give satisfactory control performance, when the controlled process is highly non-linear. The non-linearities of practical processes are normally very complex. It is quite difficult to derive complete knowledge of such non-linearities. In this paper nonlinear process such as a spherical tank system is taken as for study. The transfer function of the spherical tank system is obtained in real time by Black box modeling method. GA based FLC is developed for a level control of spherical tank process at lower region. Here the conventional FLC is designed and its servo response is compared with GA based FLC based on IAE and ISE as objective function. The proposed controller was simulated by MATLAB using the technique of PI type FLC whose membership function, fuzzy logic rules and scaling gains were optimized by the genetic algorithm technique. Simulated results show a considerable improvement in rising time and settling time besides reduces overshoot, IAE and ISE.