Intermittent sampling and detection event-based model predictive control for perturbed nonlinear systems

This paper presents a dynamic intermittent sampling strategy within event-based MPC framework for discrete-time nonlinear systems with external disturbances. A minimal triggering interval and corresponding triggering threshold are designed to treat a sub-optimal convergence property by considering the most unfavorable conditions resulted by perturbations. To reduce the conservatism in estimating the triggering interval, aperiodic sampling and detection are processed until an appropriate triggering instant is determined. In addition, a shrinking factor is incorporated to update the prediction horizon, such that the computational burden is mitigated. By applying the proposed dynamic intermittent sampling and event-based MPC, the triggering interval prolongs, such that counts of optimization decreases, and the overall computational workload is reduced. Sufficient conditions are established for recursive feasibility and stability, and simulation results demonstrate the effectiveness of the proposed scheme.