Batch-to-Batch Iterative Control Based on Multilayer Recurrent Fuzzy Neural Network for Batch Processes

Based on multilayer recurrent fuzzy neural network(MRFNN) and chaotic search(CS),the batch-to-batch iterative control strategy for final quality control in batch processes is realized.Furthermore,the strategy for temperature control in batch processes is proposed.Batch processes are modeled by MRFNN,and CS is employed for model training and optimization computation.Due to model-plant mismatches and unknown disturbances,the calculated optimal control profile may not be optimal when applied to the actual batch process.By utilizing the repetitive nature of batch process,model predictions are modified by prediction errors from previous batches,and the tracking errors are gradually reduced from batch-to-batch.The effectiveness is verified by simulation of batch reactors.