The Design of Robust Adaptive Controller Based on JITL Method for Nonlinear Process

In this paper, a robust adaptive control strategy is presented based on just-in-time learning(JITL) method for nonlinear processes. The dynamic system model of the nonlinear process can be represented by a set of local linear models based on the JITL method, and the database of JITL model is updated according to current input and output. Considering the tracking error, we introduce a feed-forward compensation which is consists of proportional and integration link to the control system, and then we can obtain a robust PID controller based on uncertain local state-space models. As a result, the simulation results show that the proposed control system has good robustness and excellent real-time control performance, and it provides a new way for the control of nonlinear processes.