Full-Text Search:
Home|Journal Papers|About CNKI|User Service|FAQ|Contact Us|中文
《Chinese Journal of Scientific Instrument》 2003-03
Add to Favorite Get Latest Update

The Research of 4-CBA Soft-sensor Model Based on BPANN

Hu Yongyou Gu Yong Su Hongye Wang Zhaohui Chu Jian (National Key Lab. of Industrial Control Technology, Institute of Advanced Process Control, Zhejiang Univers. Yuquan Campus, Hangzhou 310027,China)  
A new 4 carboxybenzaldehyde (4 CBA) soft sensor model based on three layers Back Propagation artificial neural network is developed. Some kinds of network models constructed by different neuron numbers of the hidden layer are compared in two respects:learning effect and prediction precision of test samples. Two kinds of modified BP learning algorithms are used to train the BP models with enough data samples derived from orthogonal experiments and simulation by the p Xylene oxidation kinetic model. All training and testing procedures of these BP network models are performed by programming on MATLAB. The simulation results prove the 4 CBA soft sensor model we built is far more precise and efficient than the empirical regress model (ERM) and it is possible to realize the quality control of purified terephthalic acid (PTA) product promptly in the commercial reactor.
【Fund】: 国家杰出青年科学基金 ( NSFC:60 0 2 5 3 0 8)资助项目
【CateGory Index】: TP29
Download(CAJ format) Download(PDF format)
CAJViewer7.0 supports all the CNKI file formats; AdobeReader only supports the PDF format.
©2006 Tsinghua Tongfang Knowledge Network Technology Co., Ltd.(Beijing)(TTKN) All rights reserved