Integrated Tracking Control of Batch Processes with Model Uncertainties by Updating a Linear Perturbation Model

Abstract

AbstractBased on a linear time-varying perturbation (LTVP) model, an integrated control strategy is proposed to track product quality trajectories of batch processes. To address the problem of model uncertainties occurring from batch to batch, the LTVP model is updated by using recursive discounted measurements (RDM) algorithm from the process operational data. Then batch-to-batch iterative learning control (ILC) can be feasibly combined with on-line model predictive control (MPC) within a batch. The integrated strategy can complement both methods to obtain good performance of tracking control. The proposed strategy is illustrated on a simulated batch polymerization reactor, and the results demonstrate that the performance of tracking product qualities can be improved under the proposed strategy when model uncertainties exist.