Batch-to-Batch Iterative Learning Control for End-Point Qualities Based on Kernel Principal Component Regression Model

Abstract

A batch-to-batchmodel-based iterative learning control (ILC) strategy for the end-point productquality control in batch processes is proposed in this paper. A nonlinear modelfor end-point product quality is developed from process operating data usingkernel principal component regression (KPCR). The ILC algorithm is derived tocalculate the control policy by linearizing the KPCR model around the nominaltrajectories and minimising a quadratic objective function concerning theend-point product quality. To overcome the detrimental effects of unknownprocess variations or disturbances, it is proposed in the paper that the KPCRmodel should be updated in a batchwise manner by removing the earliest batchdata from the training data set and adding the latest batch data to thetraining data set. The ILC based on updated KPCR model shows adaptability forprocess variations or disturbances when applied to a simulated batchpolymerization process. Comparisons between KPCR model and principal componentregression (PCR) model based ILCs are also made in the simulations.