Fixed-Time Composite Learning Fuzzy Control With Disturbance Rejection for Uncertain Engineering Systems Toward Industry 5.0

Intelligent control is a crucial technology for realizing Industry 5.0, which makes industrial engineering systems more efficient, robust, and resilient. It is noteworthy that uncertainties and disturbances will inevitably be detrimental to the control performances of Industry 5.0 engineering applications. To deal with these issues, we propose a novel super-twisting-like continuous extended state observer-based fixed-time composite learning fuzzy control scheme and apply it to a typical engineering system. Unlike conventional fixed-time adaptive fuzzy control methods that update parameters merely by closed-loop stability conditions, the proposed fixed-time control scheme utilizes both tracking errors and prediction errors to update parameters compositely, which achieves better-tracking performance and fuzzy approximation precision. First, fuzzy logic systems (FLSs) are developed to identify the unknown model functions in the Industry 5.0 engineering system. Second, to deal with the remaining approximation errors of the FLSs, parameter uncertainties, and external disturbances, the novel super-twisting-like continuous extended state observers are designed to estimate these lumped disturbances. Third, the prediction errors that indicate the fuzzy approximation precision are constructed by developing fixed-time parallel estimators. Moreover, rigorous Lyapunov stability analysis is carried out to illustrate the fixed-time convergence of the entire closed-loop control system. Finally, the proposed control scheme is applied to a practical buck converter engineering system toward Industry 5.0, and comparative hardware experiments verified the advantages of the control scheme.