Batch-to-batch model-based iterative optimisation control for a batch polymerisation reactor

A batch-to-batch model-based iterative optimisation control strategy for a batch polymerisation reactor is proposed. Because the reference sequences for product quality variables during a batch are usually difficult to be set reasonably and the interest usually lies in the end-of-batch product quality, a modified quadratic objective function is introduced to track the desired qualities at the end-point of a batch. Predictions of recurrent neural network models are modified using the prediction errors in the previous runs. Because model errors are gradually reduced from batch to batch, the control trajectory gradually approaches to the optimal control policy. The proposed scheme is illustrated on a simulated batch polymerisation reactor.