M-step random multiaccess protocol-based robust predictive control for Takagi-Sugeno fuzzy systems with persistent disturbances

This paper investigates the issue of M-step robust fuzzy model predictive control (RFMPC) for a category of nonlinear systems that are described using Takagi-Sugeno (T-S) fuzzy models, subject to communication network constraints and persistent bounded disturbances. To mitigate data collisions and improve data transmission reliability, a random multiaccess protocol (RMP) is integrated into the M-step RFMPC framework, resulting in the proposed M-step RMP-based RFMPC approach. Specifically, a unified representation is developed to model the interplay between T-S fuzzy nonlinearities and the RMP for nonlinear systems subject to persistent bounded disturbances. Optimization problems are formulated to determine the terminal constraint set and robust one-step sets for T-S fuzzy systems in an offline manner. Utilizing a scheduling signal-dependent quadratic function approach, a novel M-step RMP-based RFMPC algorithm is proposed, comprising both offline and online components. Sufficient conditions are established to ensure both the feasibility of the proposed algorithm and the mean-square input-to-state stability of the resulting closed-loop system. Finally, the effectiveness and applicability of the proposed M-step RMP-based RFMPC approach are demonstrated through two illustrative examples.