Title :
Progress of data-driven process monitoring for nonlinear and non-Gaussian industry process
Author :
Peiliang Wang ; Wuming He
Author_Institution :
Sch. of Inf. & Eng., Huzhou Teachers Coll., Huzhou, China
Abstract :
According to the widely existing non-Gaussian data characteristics of nonlinear processes in complex industrial process, the developments of the existing monitoring method and its application results and shortcomings are reviewed form characteristics of non-Gaussian and nonlinear, on this basis, the present situation about data-driven process monitoring and fault diagnosis of industry process with non-Gaussian and nonlinear characteristics are analyzed. The possible direction of development and the method worthy of study and the main problem to be resolved are discussed. Finally, some problems and their research tendencies in this field are presented.
Keywords :
fault diagnosis; learning (artificial intelligence); nonlinear control systems; process control; process monitoring; production engineering computing; complex industrial process; data-driven process monitoring; distribution control system; fault diagnosis; kernel-based learning method; nonGaussian data characteristics; nonGaussian industry process; nonlinear industry process; Computational modeling; Fault diagnosis; Industries; Kernel; Monitoring; Process control; Support vector machines; data-driven process monitoring; industrial Process; non-Gaussian; nonlinear;
Conference_Titel :
Information and Automation (ICIA), 2013 IEEE International Conference on
Conference_Location :
Yinchuan
DOI :
10.1109/ICInfA.2013.6720272
