Distributed SVMs based soft sensor and its application for high pressure dissolving

Author

Li, Yonggang ; Gui, Weihua ; Yang, Chunhua ; Chen, Zhisheng

Author_Institution

Sch. of Inf. Sci. & Eng., Central South Univ., Changsha

fYear

2008

fDate

25-27 June 2008

Firstpage

5611

Lastpage

5615

Abstract

High pressure dissolving (HPD) is a very important process for alumina production. During HPD process, alumina caustic ratio (ACR) of the dissolved slurry is a very important economic technical indicator. In practice, there are many factors influencing ACR and there are different noise levels for different HPD conditions. So, it is very difficult to predict ACR with single model accurately. In this paper, an improved rival penalized competitive learning clustering algorithm is used to cluster the learning samples. Then a distributed support vector machine based soft sensor is proposed to predict ACR on-line. The simulation and practical application results showed its effectiveness.

Keywords

alumina; dissolving; learning (artificial intelligence); manufacturing processes; pattern clustering; sensor fusion; slurries; support vector machines; ACR online; alumina caustic ratio; alumina production; dissolved slurry; distributed SVM; distributed support vector machine; economic technical indicator; high pressure dissolving; rival penalized competitive learning clustering algorithm; soft sensor; Analytical models; Clustering algorithms; Economic forecasting; Heating; Intelligent control; Noise level; Silicon; Slurries; Support vector machine classification; Support vector machines; Alumina caustic ratio; Distributed SVM; High pressure dissolving; Soft sensor;

fLanguage

English

Publisher

ieee

Conference_Titel

Intelligent Control and Automation, 2008. WCICA 2008. 7th World Congress on

Conference_Location

Chongqing

Print_ISBN

978-1-4244-2113-8

Electronic_ISBN

978-1-4244-2114-5

Type

conf

DOI

10.1109/WCICA.2008.4593843

Filename

4593843