Fault Diagnosis of Water Quality Monitoring Devices Based on Multiclass Support Vector Machines and Rule-Based Decision Trees
Preventing faults of sensors, wireless transmitters, and gateways are essential for water quality management in intensive aquaculture. It remains a challenging task to achieve high fault diagnostic accuracy of water quality monitoring and controlling devices. This paper proposes a hybrid water quali...
Main Authors: | , , , , |
---|---|
Format: | Article |
Language: | English |
Published: |
IEEE
2018-01-01
|
Series: | IEEE Access |
Subjects: | |
Online Access: | https://ieeexplore.ieee.org/document/8288831/ |
id |
doaj-fdba3bd6e39345a4bfc167bc99dfc1af |
---|---|
record_format |
Article |
spelling |
doaj-fdba3bd6e39345a4bfc167bc99dfc1af2021-03-29T20:54:55ZengIEEEIEEE Access2169-35362018-01-016221842219510.1109/ACCESS.2018.28005308288831Fault Diagnosis of Water Quality Monitoring Devices Based on Multiclass Support Vector Machines and Rule-Based Decision TreesShuangyin Liu0https://orcid.org/0000-0002-4068-1096Longqin Xu1Qiucheng Li2Xuehua Zhao3Daoliang Li4College of Information Science and Technology, Zhongkai University of Agriculture and Engineering, Guangzhou, ChinaCollege of Information Science and Technology, Zhongkai University of Agriculture and Engineering, Guangzhou, ChinaAdvanced System Technology (AST) Branch of Fraunhofer IOSB, Ilmenau, GermanySchool of Digital Media, Shenzhen Institute of Information Technology, Shenzhen, ChinaCollege of Information and Electrical Engineering, China Agricultural University, Beijing, ChinaPreventing faults of sensors, wireless transmitters, and gateways are essential for water quality management in intensive aquaculture. It remains a challenging task to achieve high fault diagnostic accuracy of water quality monitoring and controlling devices. This paper proposes a hybrid water quality monitoring device fault diagnosis model based on multiclass support vector machines (MSVM) in combination with rule-based decision trees (RBDT). In the modeling process, an RBDT is used to diagnose the gateway fault and wireless transmitter fault at the same time as a feature selection tool to reduce the number of features. A multiclass support vector machine classifier is employed to diagnose the faults of water quality sensors due to its robustness and generalization. We adopted an RBDT-MSVM algorithm to construct a fault diagnosis model. The diagnostic results indicate that RBDT-MSVM model has great potential for fault diagnosis of online water quality devices. RBDT-MSVM was tested and compared with other algorithms by applying it to diagnose faults of water quality monitoring devices in river crab culture ponds. The diagnostic results indicate that the model has great potential for fault diagnosis of online water quality devices. The experimental results show that the proposed model RBDT-MSVM can achieve classification accuracy as high as 92.86%, which is superior to the other three fault diagnosis methods. The results clearly confirm the superiority of the developed model in terms of classification accuracy, and that it is a suitable and effective method for fault diagnosis of water quality monitoring devices in intensive aquaculture.https://ieeexplore.ieee.org/document/8288831/Fault diagnosisMSVMrule-based decision treewireless sensor networksaquaculturewater quality |
collection |
DOAJ |
language |
English |
format |
Article |
sources |
DOAJ |
author |
Shuangyin Liu Longqin Xu Qiucheng Li Xuehua Zhao Daoliang Li |
spellingShingle |
Shuangyin Liu Longqin Xu Qiucheng Li Xuehua Zhao Daoliang Li Fault Diagnosis of Water Quality Monitoring Devices Based on Multiclass Support Vector Machines and Rule-Based Decision Trees IEEE Access Fault diagnosis MSVM rule-based decision tree wireless sensor networks aquaculture water quality |
author_facet |
Shuangyin Liu Longqin Xu Qiucheng Li Xuehua Zhao Daoliang Li |
author_sort |
Shuangyin Liu |
title |
Fault Diagnosis of Water Quality Monitoring Devices Based on Multiclass Support Vector Machines and Rule-Based Decision Trees |
title_short |
Fault Diagnosis of Water Quality Monitoring Devices Based on Multiclass Support Vector Machines and Rule-Based Decision Trees |
title_full |
Fault Diagnosis of Water Quality Monitoring Devices Based on Multiclass Support Vector Machines and Rule-Based Decision Trees |
title_fullStr |
Fault Diagnosis of Water Quality Monitoring Devices Based on Multiclass Support Vector Machines and Rule-Based Decision Trees |
title_full_unstemmed |
Fault Diagnosis of Water Quality Monitoring Devices Based on Multiclass Support Vector Machines and Rule-Based Decision Trees |
title_sort |
fault diagnosis of water quality monitoring devices based on multiclass support vector machines and rule-based decision trees |
publisher |
IEEE |
series |
IEEE Access |
issn |
2169-3536 |
publishDate |
2018-01-01 |
description |
Preventing faults of sensors, wireless transmitters, and gateways are essential for water quality management in intensive aquaculture. It remains a challenging task to achieve high fault diagnostic accuracy of water quality monitoring and controlling devices. This paper proposes a hybrid water quality monitoring device fault diagnosis model based on multiclass support vector machines (MSVM) in combination with rule-based decision trees (RBDT). In the modeling process, an RBDT is used to diagnose the gateway fault and wireless transmitter fault at the same time as a feature selection tool to reduce the number of features. A multiclass support vector machine classifier is employed to diagnose the faults of water quality sensors due to its robustness and generalization. We adopted an RBDT-MSVM algorithm to construct a fault diagnosis model. The diagnostic results indicate that RBDT-MSVM model has great potential for fault diagnosis of online water quality devices. RBDT-MSVM was tested and compared with other algorithms by applying it to diagnose faults of water quality monitoring devices in river crab culture ponds. The diagnostic results indicate that the model has great potential for fault diagnosis of online water quality devices. The experimental results show that the proposed model RBDT-MSVM can achieve classification accuracy as high as 92.86%, which is superior to the other three fault diagnosis methods. The results clearly confirm the superiority of the developed model in terms of classification accuracy, and that it is a suitable and effective method for fault diagnosis of water quality monitoring devices in intensive aquaculture. |
topic |
Fault diagnosis MSVM rule-based decision tree wireless sensor networks aquaculture water quality |
url |
https://ieeexplore.ieee.org/document/8288831/ |
work_keys_str_mv |
AT shuangyinliu faultdiagnosisofwaterqualitymonitoringdevicesbasedonmulticlasssupportvectormachinesandrulebaseddecisiontrees AT longqinxu faultdiagnosisofwaterqualitymonitoringdevicesbasedonmulticlasssupportvectormachinesandrulebaseddecisiontrees AT qiuchengli faultdiagnosisofwaterqualitymonitoringdevicesbasedonmulticlasssupportvectormachinesandrulebaseddecisiontrees AT xuehuazhao faultdiagnosisofwaterqualitymonitoringdevicesbasedonmulticlasssupportvectormachinesandrulebaseddecisiontrees AT daoliangli faultdiagnosisofwaterqualitymonitoringdevicesbasedonmulticlasssupportvectormachinesandrulebaseddecisiontrees |
_version_ |
1724194001404297216 |