Batch-to-batch iterative optimal control of batch processes based on dynamic quadratic criterion

Abstract

A novel dynamic parameters-based quadratic criterion-iterative learning control is proposed in this paper. Firstly, quadratic criterion-iterative learning control with dynamic parameters is used to improve the performance of iterative learning control. As a result, the proposed method can avoid the problem of initialization of the optimization controller parameters, which are usually resorted to trial and error procedure in the existing iterative algorithms used for the optimization of batch process. Next, we make the first attempt to give rigorous description and proof to verify that a perfect tracking performance can be obtained. Lastly, examples are used to illustrate the performance and applicability of the proposed method. © 2010 Springer-Verlag.