Self-triggered model predictive control for nonlinear continuous-time networked system via ensured performance control samples selection

Abstract

Self-triggered and event-triggered control have been widely implemented in various fields recently, especially in networked control systems. In order to save communication resources, trigger control conducts aperiodic control strategies instead of traditional time-driven scheme. Based on the self-triggered control mechanism, this paper investigates self-triggered strategy in model predictive control for nonlinear continuous-time networked system, selects aperiodic samples to ensure system performance while reducing computing energies and improving the resource usage efficiency. The usage of communication resources is quantified as a damping parameter that can characterize the communication effects in total cost function, such that the optimal sampling time and input trajectory can be obtained synchronously. In theoretical results, the feasibility of the proposed self-triggered strategy and the asymptotic stability of the nonlinear system are guaranteed. Furthermore, a simulation is given to illustrate the effectiveness of the proposed approach.