Multibatch Model Predictive Control for Repetitive Batch Operation with Input-Output Linearization

Junghui Chen

doi:10.1021/ie2020125

Return

Multibatch Model Predictive Control for Repetitive Batch Operation with Input-Output Linearization

2012-07-03

PRE

Junghui Chen *

Yu-Hung Lin

DOI：10.1021/ie2020125

Original

Original request for help

Save

Abstract

En 中文

Because of the repetitive nature of the batch process, in conventional batch process control, the start of the subsequent batch is often assumed to be the same as that of the prior batch. Nevertheless, in real operation, under the influence of various exterior factors, most batch processes involve gradual changes over batches and, thus, the end-point condition would affect the start of the subsequent process batch. The initial condition at each batch may not be reset to be the desired initial condition. In order to ensure an optimum operation, it is essential to taking into account the effect of the previous batch. In this paper, a multibatch dynamics model is derived. It takes into account this batchwise dynamics. The derived process model can provide the benefits of better prediction and performance. In order to overcome the problem of nonlinearity in the batch process, the nonlinear process is transformed to a linear process through input output feedback linearization. The identification method of the multibatch model is proposed for such batch processes, and the model will be able to reflect the particular type of batch process, which can be transformed to a linear input output relation. Based on this model, a model predictive control scheme is derived for the case of unconstrained control as well as constrained control. The applications are discussed through a fed-batch reactor problem for the biosynthesis of penicillin to demonstrate the advantages of the proposed method, in comparison with the conventional methods.

Keywords:

ITERATIVE LEARNING CONTROL

FEEDBACK

AI Summary

Key information extracted from the uploaded paper, including a brief overview, abstract, background, key highlights, visual analysis, and future outlook.