Quantized Robust Event-Triggered MPC for Linear Discrete-Time Systems

This paper presents a Quantized Event-Triggered Model Predictive Control (QET-MPC) framework for linear discrete-time systems with bounded disturbances, introducing a unified treatment of event-triggering and dynamic quantization within a dual-mode MPC architecture. Unlike conventional event-triggered MPC designs, the proposed approach explicitly accounts for the effects of quantization on both stability and feasibility, which have not been analyzed before. The event-triggering condition is defined over quantized state deviations, enabling the MPC optimization to be solved only when the quantized actual and predicted trajectories diverge beyond a specified threshold—thereby reducing computational effort. A dynamic quantizer with a “zoom-in/zoom-out” protocol adjusts its precision online according to the system’s state magnitude, allowing the quantization precision to increase near the equilibrium. Rigorous recursive feasibility is established and stability guarantees are derived: Lyapunov-based analysis establishes cost decrease when the quantized states are outside the terminal region, while input-to-state stability (ISS) is proven once the states enter the quantized feedback mode. Simulation results confirm that the proposed QET-MPC maintains stability and constraint satisfaction despite quantization effects, achieving high control accuracy with significantly fewer MPC updates compared to conventional time-triggered designs. Finally, the influence of quantization parameters on closed-loop performance is analyzed, highlighting their role in shaping precision and control efficiency.