Fixed‐Time Convergence Adaptive Robust Controller for Overhead Cranes Based on Reinforcement Learning Considering Input Saturation

• Published in: International journal of adaptive control and signal processing Vol. 39; no. 6; pp. 1252 - 1273
• Main Authors: Wei, Junren, Xu, Weimin
• Format: Journal Article
• Language: English
• Published: Bognor Regis Wiley Subscription Services, Inc 01.06.2025
• Subjects: Control methods; Control systems; Controllers; Convergence; Cranes; Disturbances; Error compensation; Feedforward control; Machine learning; Parameters; Robust control; Sliding mode control; Stability analysis; State observers
• ISSN: 0890-6327; 1099-1115
• DOI: 10.1002/acs.4005

An adaptive fixed‐time extended state observer (AFESO)‐based nonsingular fast terminal sliding mode control (NFTSMC) method is proposed for a class of underactuated overhead crane systems with unknown external disturbances. Firstly, an improved fixed‐time nonsingular fast terminal sliding mode surface is designed, which not only solves the singularity problems in terminal sliding mode control but also achieves system stabilization within bounded convergence time. Moreover, an adaptive fixed‐time extended state observer is presented to estimate both matched and unmatched disturbances; then, the estimated value is used for feedforward compensation to compose the controller. Subsequently, to improve control quality and enhance robustness of the control system, a weakly reinforcement learning (RL) algorithm is used to update parameters of a chattering free reaching law parameter, upon which a novel fixed‐time auxiliary error compensation system is constructed to solve the input saturation problem. The fixed‐time stability analysis of the closed‐loop system is derived by Lyapunov theory. Finally, the rapidity and effectiveness of the developed control method are demonstrated through simulation results.