An integrated batch-to-batch iterative learning control and within batch control strategy for batch processes

Abstract

An integrated batch-to-batch iterative learning control (ILC) and within batch on-line shrinking horizon model predictive control (SHMPC) strategy for the tracking control of product qualities in batch processes is proposed. ILC is used for batch-to-batch control based on a batch-wise linear time-varying (LTV) perturbation model and the convergence of batch-wise tracking error under ILC is guaranteed. On-line SHMPC within a batch can reduce the effects of disturbances immediately and improve the performance of the current batch run. The on-line model prediction can be also obtained based on the batch-wise LTV model. The integrated control strategy can complement both methods to obtain good performance of tracking control. The proposed strategy is illustrated on a simulated batch polymerization process.