Stability of Constrained Model Predictive Control for Networked Control Systems with Data Packet Dropout

A constrained model predictive control (MPC) strategy for networked control systems with data packet dropout is proposed in this paper. A buffer is designed to store the predicted control sequence between controller and actuator. If data packet dropout doesn't happen, new control sequence will be obtained by MPC to substitute old control sequence in buffer and the first control action will be enforced to the actuator, or the control action in old control sequence corresponding to current sample time will be applied. It is also shown that if the control horizon of MPC is not less than maximal continuous data packet dropout number, feasibility of MPC at initial time implies asymptotical stability of the closed loop system. A simulation example illustrates the effectiveness of the proposed approach.