Fuzzy Optimization and Control for A Non-linear Process using Extended Kalman Filter

摘要

The proposed fuzzy controller with Extended Kalman Filter is designed to control the performance of nonlinear process. The nonlinear process taken up for the study is Two Tank Conical Non Interacting System. Control of flow in a conical tank is important because the shape of the liquid in the tank varies continuously which gives rise to the nonlinearity. The mathematical model of the two tank conical system is developed with two inputs and one output. To achieve the level control in the conical tank system, fuzzy membership function optimization is to be carried out by using Extended Kalman Filter (EKF) which is the popular optimization tool. EKF resembles the predictor corrector algorithm which is used to optimize the input membership function of the fuzzy controller to get a desired output. Fuzzy optimization using EKF is compared with the conventional PID controller.