Event-triggered distributed model predictive control for constrained linear system with random packet loss

This paper investigates the problem of event-triggered distributed model predictive control (MPC) for a set of linear local systems subject to random packet loss. We present a new model for each local system associated with a cost involving neighboring system states, in which a stochastic variable is adopted to describe successive packet loss in a random way. To achieve the stability of the global closed-loop system, a compatibility condition is imposed in the distributed MPC optimization problem. Moreover, an event-triggered communication scheme is constructed to determine whether the optimization problem is solved or not, which reduces the communication energy consumption and lowers the computational burden. Based on the linear matrix inequalities (LMIs) technique, we propose an algorithm by which the closed-loop stability of the whole system is guaranteed. In final, numerical example is given to illustrate the effectiveness of the proposed approach.