Title :
A Fault Diagnosis Method Based on Composite Model and SVM for Fermentation Process
Author :
Ma, Liling ; Wang, Junzheng ; Liu, Zhigang
Author_Institution :
Beijing Inst. of Technol., Beijing
fDate :
May 30 2007-June 1 2007
Abstract :
A method of fault diagnosis based on composite model and support vector machines for fermentation process is proposed to overcome its difficulty in direct measurement of state parameters. In order to obtain the process state, composite model is presented by combining mass equations of bioreactors with RBF neural network that serve as estimators of unmeasured process kinetic parameters. Then Support vector machines are used to analyze and recognize fault patterns, making use of estimated state variables on line. The proposed method is applied to glutamic acid fermentation process, and the simulation results show its feasibility and effectiveness.
Keywords :
bioreactors; fault diagnosis; fermentation; production engineering computing; radial basis function networks; support vector machines; RBF neural network; SVM; bioreactors; fault diagnosis method; fault patterns; glutamic acid fermentation process; support vector machines; Amino acids; Bioreactors; Equations; Fault diagnosis; Kinetic theory; Neural networks; Pattern analysis; Pattern recognition; State estimation; Support vector machines; composite model; fault diagnosis; fermentation process; support vector machine;
Conference_Titel :
Control and Automation, 2007. ICCA 2007. IEEE International Conference on
Conference_Location :
Guangzhou
Print_ISBN :
978-1-4244-0818-4
Electronic_ISBN :
978-1-4244-0818-4
DOI :
10.1109/ICCA.2007.4376933
