A gain scheduled PID control method based on nonlinear model predictive control

Abstract

This paper tries to design self-tuning PID control systems for non-linear systems, which optimize a certain performance index. The key idea of this research is to utilize non-linear model predictive control method in order to design the PID gains. After designing the closed loop systems by using PID controller, the PID gains are regarded as manipulated variables. Then, non-linear model predictive control systems are designed where the performance index consists of integrated quadratic forms of tracking error and manipulated variables representing PID gains. The proposed controller is categorized as gain-scheduled controller because the PID gains are tuned using state variables as scheduling parameter. That is why we call it the gain scheduled PID controller. The effectiveness of the proposed method is shown by comparative simulation results with the fixed gain PID control systems.