Observer Based Model-Free Adaptive Iterative Learning Constrained Control for Nonlinear Systems

In this paper, the large class of nonlinear control systems with repeating tasks are addressed by the proposal of a new model-free adaptive iterative learning (MFAILC) constrained control strategy. With the aid of the compact form dynamic linearization (CFDL) technique, a new observer-based pseudo partial derivative (PPD) iterative estimation algorithm is created. Then, an anti-windup compensator would be suggested to modify reference trajectory in to avoid parameter expansion and system instability, with the goal of solving the input constraint problem driven on by actuator saturation. Furthermore, an iterative constrained controller is proposed and the stability of the controller is proved. Finally, it is demonstrated by numerical simulation that the suggested control algorithm has excellent tracking capability and reliability.