Robust Soft Sensor with Deep Kernel Learning for Quality Prediction in Rubber Mixing Processes
Although several data-driven soft sensors are available, online reliable prediction of the Mooney viscosity in industrial rubber mixing processes is still a challenging task. A robust semi-supervised soft sensor, called ensemble deep correntropy kernel regression (EDCKR), is proposed. It integrates...
Main Authors: | , , , , , |
---|---|
Format: | Article |
Language: | English |
Published: |
MDPI AG
2020-01-01
|
Series: | Sensors |
Subjects: | |
Online Access: | https://www.mdpi.com/1424-8220/20/3/695 |
id |
doaj-e5c9307029e74f47bc20d13d780f171e |
---|---|
record_format |
Article |
spelling |
doaj-e5c9307029e74f47bc20d13d780f171e2020-11-25T01:12:57ZengMDPI AGSensors1424-82202020-01-0120369510.3390/s20030695s20030695Robust Soft Sensor with Deep Kernel Learning for Quality Prediction in Rubber Mixing ProcessesShuihua Zheng0Kaixin Liu1Yili Xu2Hao Chen3Xuelei Zhang4Yi Liu5Institute of Process Equipment and Control Engineering, Zhejiang University of Technology, Hangzhou 310023, ChinaInstitute of Process Equipment and Control Engineering, Zhejiang University of Technology, Hangzhou 310023, ChinaShanghai Customs, Shanghai 200120, ChinaQuanzhou Institute of Equipment Manufacturing, Haixi Institutes, Chinese Academy of Sciences, Jinjiang 362200, ChinaShanghai Customs, Shanghai 200120, ChinaInstitute of Process Equipment and Control Engineering, Zhejiang University of Technology, Hangzhou 310023, ChinaAlthough several data-driven soft sensors are available, online reliable prediction of the Mooney viscosity in industrial rubber mixing processes is still a challenging task. A robust semi-supervised soft sensor, called ensemble deep correntropy kernel regression (EDCKR), is proposed. It integrates the ensemble strategy, deep brief network (DBN), and correntropy kernel regression (CKR) into a unified soft sensing framework. The multilevel DBN-based unsupervised learning stage extracts useful information from all secondary variables. Sequentially, a supervised CKR model is built to explore the relationship between the extracted features and the Mooney viscosity values. Without cumbersome preprocessing steps, the negative effects of outliers are reduced using the CKR-based robust nonlinear estimator. With the help of ensemble strategy, more reliable prediction results are further obtained. An industrial case validates the practicality and reliability of EDCKR.https://www.mdpi.com/1424-8220/20/3/695soft sensordeep learningsemi-supervised learningrobust estimatorensemble strategyrubber mixing processmooney viscosity |
collection |
DOAJ |
language |
English |
format |
Article |
sources |
DOAJ |
author |
Shuihua Zheng Kaixin Liu Yili Xu Hao Chen Xuelei Zhang Yi Liu |
spellingShingle |
Shuihua Zheng Kaixin Liu Yili Xu Hao Chen Xuelei Zhang Yi Liu Robust Soft Sensor with Deep Kernel Learning for Quality Prediction in Rubber Mixing Processes Sensors soft sensor deep learning semi-supervised learning robust estimator ensemble strategy rubber mixing process mooney viscosity |
author_facet |
Shuihua Zheng Kaixin Liu Yili Xu Hao Chen Xuelei Zhang Yi Liu |
author_sort |
Shuihua Zheng |
title |
Robust Soft Sensor with Deep Kernel Learning for Quality Prediction in Rubber Mixing Processes |
title_short |
Robust Soft Sensor with Deep Kernel Learning for Quality Prediction in Rubber Mixing Processes |
title_full |
Robust Soft Sensor with Deep Kernel Learning for Quality Prediction in Rubber Mixing Processes |
title_fullStr |
Robust Soft Sensor with Deep Kernel Learning for Quality Prediction in Rubber Mixing Processes |
title_full_unstemmed |
Robust Soft Sensor with Deep Kernel Learning for Quality Prediction in Rubber Mixing Processes |
title_sort |
robust soft sensor with deep kernel learning for quality prediction in rubber mixing processes |
publisher |
MDPI AG |
series |
Sensors |
issn |
1424-8220 |
publishDate |
2020-01-01 |
description |
Although several data-driven soft sensors are available, online reliable prediction of the Mooney viscosity in industrial rubber mixing processes is still a challenging task. A robust semi-supervised soft sensor, called ensemble deep correntropy kernel regression (EDCKR), is proposed. It integrates the ensemble strategy, deep brief network (DBN), and correntropy kernel regression (CKR) into a unified soft sensing framework. The multilevel DBN-based unsupervised learning stage extracts useful information from all secondary variables. Sequentially, a supervised CKR model is built to explore the relationship between the extracted features and the Mooney viscosity values. Without cumbersome preprocessing steps, the negative effects of outliers are reduced using the CKR-based robust nonlinear estimator. With the help of ensemble strategy, more reliable prediction results are further obtained. An industrial case validates the practicality and reliability of EDCKR. |
topic |
soft sensor deep learning semi-supervised learning robust estimator ensemble strategy rubber mixing process mooney viscosity |
url |
https://www.mdpi.com/1424-8220/20/3/695 |
work_keys_str_mv |
AT shuihuazheng robustsoftsensorwithdeepkernellearningforqualitypredictioninrubbermixingprocesses AT kaixinliu robustsoftsensorwithdeepkernellearningforqualitypredictioninrubbermixingprocesses AT yilixu robustsoftsensorwithdeepkernellearningforqualitypredictioninrubbermixingprocesses AT haochen robustsoftsensorwithdeepkernellearningforqualitypredictioninrubbermixingprocesses AT xueleizhang robustsoftsensorwithdeepkernellearningforqualitypredictioninrubbermixingprocesses AT yiliu robustsoftsensorwithdeepkernellearningforqualitypredictioninrubbermixingprocesses |
_version_ |
1725164093674881024 |