Low-Complexity High-Performance Control of Unknown Block-Triangular MIMO Nonlinear Systems

Abstract

This article focuses on the high-performance tracking control problem for the unknown multi-input multi-output (MIMO) nonlinear systems in the block triangular form with unmatched disturbances. It aims to not only accomplish prescribed performance tracking, but also actively suppress error oscillations arising from unmodelled dynamics, persistent disturbances, and time-varying references. The existing methods are applicable to the first-order systems or the single-input single-output systems, or necessitate the available system nonlinearities or reference derivatives, or invoke approximating structures and adaptive mechanisms. In this article, a novel proportional-integral (PI) prescribed performance control (PPC) approach is put forward, which consists of a mixed-gain adaptation technique and a smooth switching PPC scheme. The controller is independent of the specific model information of the plant or the derivatives of the references and the intermediate control signals. Furthermore, it maintains the distinctive simplicity of both PI control and PPC, with no parameter identification, function approximation, disturbance estimation, or command filtering. The effectiveness and superiority of our approach are illustrated through a comparative experiment on a 2-DOF serial flexible link robot.