DocumentCode :
175407
Title :
Soft sensor modeling of mill level based on Deep Belief Network
Author :
Muchao Lu ; Yan Kang ; Xiaoming Han ; Gaowei Yan
Author_Institution :
Coll. of Inf. Eng., Taiyuan Univ. of Technol., Taiyuan, China
fYear :
2014
fDate :
May 31 2014-June 2 2014
Firstpage :
189
Lastpage :
193
Abstract :
Accurate measurement of the mill level is a key factor to improve the ball mill´s productive efficiency, safety and economy. Aiming at solving the critical problem of the mill level soft sensor, feature extraction of the processing parameters, a novel method based on Deep Belief Network (DBN) is proposed. DBN is one of the deep learning methods, which focuses on learning deep hierarchical models of data. In this paper, basic features, namely power spectrum density are obtained from the vibration signal of ball mill by Welch´s method firstly. Then DBN is built on the basic features to learn high level deep features. Finally a supervised learning algorithm named back propagation neural network is used to model the relationships between extracted features and mill level. Experimental results indicate that the DBN based method outperforms traditional feature extraction methods.
Keywords :
backpropagation; ball milling; belief networks; feature extraction; milling machines; neural nets; production engineering computing; vibrations; DBN; back propagation neural network; ball mill economy; ball mill productive efficiency; ball mill safety; ball mill vibration signal; deep belief network; deep hierarchical data model learning; deep learning methods; mill level soft sensor; power spectrum density; processing parameter feature extraction; soft sensor modeling; supervised learning algorithm; Computational modeling; Feature extraction; Indexes; Measurement; Monitoring; Principal component analysis; Vibrations; back propagation neural network; deep belief network; feature extraction; mill level;
fLanguage :
English
Publisher :
ieee
Conference_Titel :
Control and Decision Conference (2014 CCDC), The 26th Chinese
Conference_Location :
Changsha
Print_ISBN :
978-1-4799-3707-3
Type :
conf
DOI :
10.1109/CCDC.2014.6852142
Filename :
6852142
Link To Document :
بازگشت