Dual Observer-Based Model-Free Adaptive I/O Constrained Control for MIMO Nonlinear Systems

Abstract

In this paper, a dual observer based model-free adaptive control strategy is designed for multiple input multiple output (MIMO) nonlinear systems with disturbances and input/output (I/O) constraints. The dual observers consists of an adaptive observer and a discrete extended state observer, in which the former is designed to realize the dynamic reconfiguration of the system and devise the Lyapunov stability criterion-based estimation algorithm for time-varying parameters, and the latter is explored for composite disturbance estimation. Based on the information from dual observers, a dynamic anti-windup compensator along with an improved prescribed performance control method are proposed in the sliding mode controller to solve the I/O constraint problem. Finally, the stability analysis and simulation are supplied for performance verification.