Error Expectation Based Dynamic Event-Triggered Stochastic MPC with Adaptive Prediction Horizon*

A dynamic event-triggered stochastic model predictive control (SMPC) algorithm is proposed for discrete-time systems subject to stochastic uncertainties and probabilistic constraints. The proposed scheme optimizes both the event-triggering frequency and computational load, effectively reducing the resource consumption associated with SMPC. In this framework, the dynamic event-triggered mechanism incorporates statistical characteristics, such as error expectations, to set the triggering conditions. Additionally, the prediction horizon can dynamically adjust, reducing computational complexity of the optimization problem. Theoretical results ensure recursive feasibility and establish conditions for quadratic stability. Finally, the simulation results confirm the effectiveness of the developed dynamic event-triggered stochastic control algorithm.