Adaptive Finite-Time Neural Control for Uncertain Markov Jump Systems with Actuator Faults

This article delves into the intricate world of Markov jump systems (MJSs), hybrid dynamic systems that switch between multiple subsystems via a Markov mechanism. The focus is on adaptive finite-time neural control, a cutting-edge approach to managing uncertainties and actuator faults in these systems. The article highlights the advantages of finite-time stability, which offers faster convergence and better robustness compared to traditional asymptotic convergence. It introduces a novel sliding mode control (SMC) law based on adaptive radial basis function neural networks (RBFNN), which can handle unknown disturbances and composite actuator faults without relying on restrictive assumptions about fault boundaries. The article presents a special linear sliding mode surface and partitioning strategy to derive new sufficient conditions for finite-time boundedness, ensuring the system state remains within a certain threshold during a specified finite-time interval. Through rigorous analysis and simulation, the article demonstrates the effectiveness of the proposed approach in achieving fast convergence and high precision performance, even in the presence of unknown actuator faults and disturbances. The simulations involve an F-404 aircraft engine system, showcasing the practical applicability of the method. The article also includes a quantitative evaluation, comparing the proposed scheme with existing methods to highlight its superior control accuracy and lower energy consumption.