A Novel Nonlinear Model Predictive Control Strategy for a Class of Nonlinear Systems with Multiple Actuators’ Response Time-Delays

<div class="section abstract"><div class="htmlview paragraph">This paper investigates the problem of nonlinear model predictive control (NMPC) strategy for a class of nonlinear systems with multiple actuators’ response time-delays. Conventional approaches that incorporate these time-delays into the NMPC formulation typically result in a significant increase in the optimization problem's scale. To address these problems, we propose a novel NMPC strategy. In the first stage, the NMPC strategy is designed for the nonlinear system without considering actuator’s response time-delay, thereby maintaining the original scale of the optimization problem. The optimal control sequence derived from this NMPC is then fitted to a time-continuous polynomial function, serving as a reference signal for the actuators' response time-delay models. In the second stage, combining inverse model and inverse Laplace transform techniques, a novel inverse model compensation control (IMCC) strategy is designed for actuators’ response time-delays. This IMCC strategy enables tracking of the reference signal without phase time-delay or amplitude deviation. For comparative analysis, we also implement a model augmentation NMPC strategy that directly incorporates actuators’ time-delays, inevitably increasing the scale of the optimization problem. By quantitative analysis, the model augmentation NMPC strategy will increase the number of optimal variables and equality constraints of the optimization problem. Finally, vehicle control of transport vehicle in open-pit mine is taken as simulation example, the simulation results show that both the proposed novel NMPC and IMCC algorithms and model augmentation NMPC algorithm can achieve high precision control performance, the maximum and average calculation time of the proposed novel NMPC and IMCC algorithms are 31.9% and 46.2% lower than that of model augmentation NMPC algorithm, respectively.</div></div>