Input fluctuation mitigation in coupled nonlinear control systems: A dual-mode DMPC approach

Abstract

Excessive fluctuations in control inputs can severely compromise the performance, stability, and safety of control systems, posing significant challenges in various practical applications. This study introduces a novel dual-mode distributed model predictive control (DMPC) approach for a class of dynamically coupled nonlinear systems, aiming to effectively mitigate input fluctuations. The existence of a terminal invariant region (TIR) and its corresponding terminal stabilizing controller (TSC) is rigorously established for a globally coupled nonlinear system subject to input amplitude and fluctuation constraints. When the system state resides within the global TIR, the global TSC ensures consistent satisfaction of input amplitude and fluctuation constraints. When operating outside the global TIR, a set of parallel model predictive controllers with input amplitude and fluctuation constraints is developed, where an aperiodic event-triggering scheduler is introduced to reduce the likelihood of input fluctuation constraint violations during practical numerical implementation. The recursive feasibility and closed-loop stability of the proposed dual-mode DMPC approach are rigorously analyzed. Simulation results on coupled oscillators demonstrate its effectiveness in input fluctuation mitigation and system stabilization.