Model Predictive Control with Variable Predictive Horizon for Remote Control System including Variable Delay

抄録

In this paper, a model predictive control (MPC) based time-varying delay compensation system is proposed. The proposed method uses a prediction model that estimates the future state of a remote plant to compensate for the delay in remote control. Because the amount of the delay is not known before transmission, the local prediction model needs to assume the maximum delay when the delay varies. However, most packets arrive within the maximum delay time. Therefore, this study proposes the utilization of early-arrival packet to generate a reference trajectory for model predictive control at the remote side. To achieve this, we propose a method to deal with the variable horizon for model predictive control. In addition, the proposed method predicts the state of the remote plant considering the controller on the remote side; therefore, it is possible to predict the state with a higher accuracy than the conventional method. The performance of this method in comparison with the conventional method is evaluated by numerical simulations and experiments.