Iterative Learning Control and It´s Application to Batch Process Optimization

Author

Song, J-R ; Wang, H-W ; Shi, H-B ; Zhang, SH-H

Author_Institution

Sch. of Inf. Sci. & Eng., East China Univ. of Sci. & Technol., Shanghai, China

fYear

2011

fDate

17-18 July 2011

Firstpage

1

Lastpage

4

Abstract

An iterative learning control (ILC) algorithm based on recurrent wavelet neural network(RWNN) is proposed to control product final quality in batch process. recurrent Wavelet neural network is used to modeling long range batch process model. Due to model-plant mismatches and unmeasured disturbances, the calculated control policy based on RWNN model may not be optimal when applied to the actual process. By utilizing the repetitive nature of batch process , ILC is used to improve product final quality from batch to batch. Prediction models are modified based on previous prediction model average errors. Model errors are gradually reduced from batch to batch, control inputs approach to optimal control policy. The effectiveness is verified on a simulated batch process.

Keywords

batch processing (industrial); iterative methods; learning (artificial intelligence); optimal control; optimisation; process control; production control; quality control; recurrent neural nets; wavelet transforms; ILC; RWNN model; batch process model; batch process optimization; iterative learning control algorithm; optimal control policy; prediction model average error; product final quality control; recurrent wavelet neural network; Batch production systems; Indexes; Optimal control; Predictive models; Process control; Recurrent neural networks; Trajectory;

fLanguage

English

Publisher

ieee

Conference_Titel

Circuits, Communications and System (PACCS), 2011 Third Pacific-Asia Conference on

Conference_Location

Wuhan

Print_ISBN

978-1-4577-0855-8

Type

conf

DOI

10.1109/PACCS.2011.5990241

Filename

5990241