Two-Dimensional Iterative Learning Robust Asynchronous Switching Predictive Control for Multiphase Batch Processes With Time-Varying Delays

This study formulated an iterative learning-based predictive control strategy for asynchronous switching of multiphase batch processes with complex characteristics in the framework of a two-dimensional (2-D) system. First, we constructed a Fornasini-Marchesini comprehensive feedback error model, considering the state deviation and output error. Using this model, we developed a switching model considering the match and mismatch cases. Furthermore, an iterative learning-based predictive control mechanism was designed for asynchronous switching with a greater freedom of adjustment and fast learning ability in the batch direction. Second, the asymptotic and exponential stability were discussed based on the related methods and theories, and the system stability conditions were expressed in the form of linear matrix inequality (LMI). Following an online mechanism to determine the LMI conditions, we derived the real-time optimal gains of the control law, the maximum dwell period (Max-DT) for the mismatch case, and the minimum dwell period for the match case. The switching signal was transmitted in advance according to the Max-DT to ensure the stability of the system during switching. Finally, the effectiveness of the proposed method was confirmed by utilizing the injection molding process.