Design of Fuzzy Adaptive Pi Controller for Inherently Unstable System

Abstract

Desired performance of any process depends on the control action taken by the controller. Choice of control structure and the design approach is mainly concerned with a priori knowledge of the process dynamics governed by any system, its limitations to achieve desire response and calculation of system state at different instant of time. Many practitioners and researchers in the field of control agree that controller design can be quite laborious in cases where the plant is unstable. The controller synthesis becomes a challenge for such systems because there are certain design and performance limitations that narrow the range of feasible solutions. These constraints reflect imperfect controller design which may work satisfactorily in one situation and may fail completely in other. Adaptive control where the controller parameters adjust them self as per the system requirements, is a candidate solution for such problem. This paper mainly focuses on design of model reference adaptive control based PI (Proportional-Integral) controller for unstable system. Adaptation mechanism is developed by fuzzy logic which makes the controller more versatile and reduces the design requirements. Results are compared with the conventional controller tuned with Ziegler-Nichols method. Detailed simulation Study and experimental validation shows the effectiveness of proposed Fuzzy Adaptive PI controller to control the unstable system for obtaining the desired performance out of it.