Neural Network Functional Observer-Based Composite Anti-Disturbance Control for Systems With Multiplicative and Implicit Disturbances

Abstract

High precision control under disturbances and uncertainties is critical to the safe, stable, and long-term continuous operation of the system. In this paper, a composite anti-disturbance control strategy for systems affected by both multiplicative and implicit disturbances is proposed. First, a reduced neural network functional observer is developed to estimate the partially unknown states and the implicit disturbances, which takes into account the effect of the multiplicative disturbances and the uncertainties in the implicit disturbances, both of which are related to the system's states. The neural network is employed for approximating the multiplicative disturbances. Then, a dynamic sliding mode surface with disturbance compensation is introduced to ensure exponential convergence of systems with the multiplicative disturbances and the implicit disturbances. Furthermore, a novel barrier function-based adaptive sliding mode control law is designed to guarantee that the system trajectories reach the sliding surface, which greatly reduces chattering without requiring prior knowledge of the upper bound on system uncertainties. Finally, a numerical example is presented to demonstrate the effectiveness of the proposed approach.