Title :
Multi-scale Wavelet Support Vector Regression for Soft Sensor Modeling
Author :
Wang, Jun ; Peng, Hong
Author_Institution :
Sch. of Electr. Inf., Xihua Univ., Sichuan
Abstract :
A new multi-scale wavelet support vector regression (MS-WSVR) for function regression is proposed in this paper, and applied to soft sensor modeling for abamectin fermentation process. Wavelet function with different resolution is used as support vector kernel in order to construct MS-WSVR. Theoretic analysis of the multi-scale wavelet kernel is discussed in detail. Experiments show that this method has better performance than the soft sensor modeling based on the RBF neural network
Keywords :
chemical engineering computing; fermentation; neural nets; support vector machines; wavelet transforms; abamectin fermentation process; function regression; multi-scale wavelet support vector regression; soft sensor modeling; support vector kernel; Computer science; Kernel; Mathematical model; Mathematics; Neural networks; Neurofeedback; Process control; Support vector machines; Training data; Wavelet analysis;
Conference_Titel :
Neural Networks and Brain, 2005. ICNN&B '05. International Conference on
Conference_Location :
Beijing
Print_ISBN :
0-7803-9422-4
DOI :
10.1109/ICNNB.2005.1614616