Sub-Health Identification of Reciprocating Machinery Based on Sound Feature and OOD Detection
It is inevitable that machine parts will be worn down in production, causing other mechanical failures. With the appearance of wearing, the accuracy and efficiency of machinery gradually decline. The state between healthy and impaired is defined as sub-health. By recognizing the sub-health state of...
Main Authors: | , , , |
---|---|
Format: | Article |
Language: | English |
Published: |
MDPI AG
2021-08-01
|
Series: | Machines |
Subjects: | |
Online Access: | https://www.mdpi.com/2075-1702/9/8/179 |
id |
doaj-28e9060cd2a84a919c69b81ba30a8f3a |
---|---|
record_format |
Article |
spelling |
doaj-28e9060cd2a84a919c69b81ba30a8f3a2021-08-26T13:59:39ZengMDPI AGMachines2075-17022021-08-01917917910.3390/machines9080179Sub-Health Identification of Reciprocating Machinery Based on Sound Feature and OOD DetectionPeng Cui0Jinjia Wang1Xiaobang Li2Chunfeng Li3School of Information Science and Engineer, Yanshan University, Qinhuangdao 066004, ChinaSchool of Information Science and Engineer, Yanshan University, Qinhuangdao 066004, ChinaIndustrial Technology Center, Hebei Petroleum University of Technology, Chengde 067000, ChinaIndustrial Technology Center, Hebei Petroleum University of Technology, Chengde 067000, ChinaIt is inevitable that machine parts will be worn down in production, causing other mechanical failures. With the appearance of wearing, the accuracy and efficiency of machinery gradually decline. The state between healthy and impaired is defined as sub-health. By recognizing the sub-health state of machinery, accuracy and efficiency can be effectively guaranteed, and the occurrence of mechanical failure can be prevented. Compared with simple fault detection, the identification of s sub-health state has more practical significance. For this reason, the sound characteristics of large-scale reciprocating machinery, combined with the concept of OOD (out-of-distribution) detection, are used, and a model for detecting machinery sub-health state is proposed. A planer sound dataset was collected and collated, and the recognition of mechanical sub-health state was realized by a model combining a VGG network and the threshold setting scheme of OOD detection. Finally, an auxiliary decision-making module was added, and Mahalanobis distance was used to represent spatial relationships among samples, further improving the recognition effect.https://www.mdpi.com/2075-1702/9/8/179fault detectionout-of-distributionconvolutional neural networksound featurestate detection |
collection |
DOAJ |
language |
English |
format |
Article |
sources |
DOAJ |
author |
Peng Cui Jinjia Wang Xiaobang Li Chunfeng Li |
spellingShingle |
Peng Cui Jinjia Wang Xiaobang Li Chunfeng Li Sub-Health Identification of Reciprocating Machinery Based on Sound Feature and OOD Detection Machines fault detection out-of-distribution convolutional neural network sound feature state detection |
author_facet |
Peng Cui Jinjia Wang Xiaobang Li Chunfeng Li |
author_sort |
Peng Cui |
title |
Sub-Health Identification of Reciprocating Machinery Based on Sound Feature and OOD Detection |
title_short |
Sub-Health Identification of Reciprocating Machinery Based on Sound Feature and OOD Detection |
title_full |
Sub-Health Identification of Reciprocating Machinery Based on Sound Feature and OOD Detection |
title_fullStr |
Sub-Health Identification of Reciprocating Machinery Based on Sound Feature and OOD Detection |
title_full_unstemmed |
Sub-Health Identification of Reciprocating Machinery Based on Sound Feature and OOD Detection |
title_sort |
sub-health identification of reciprocating machinery based on sound feature and ood detection |
publisher |
MDPI AG |
series |
Machines |
issn |
2075-1702 |
publishDate |
2021-08-01 |
description |
It is inevitable that machine parts will be worn down in production, causing other mechanical failures. With the appearance of wearing, the accuracy and efficiency of machinery gradually decline. The state between healthy and impaired is defined as sub-health. By recognizing the sub-health state of machinery, accuracy and efficiency can be effectively guaranteed, and the occurrence of mechanical failure can be prevented. Compared with simple fault detection, the identification of s sub-health state has more practical significance. For this reason, the sound characteristics of large-scale reciprocating machinery, combined with the concept of OOD (out-of-distribution) detection, are used, and a model for detecting machinery sub-health state is proposed. A planer sound dataset was collected and collated, and the recognition of mechanical sub-health state was realized by a model combining a VGG network and the threshold setting scheme of OOD detection. Finally, an auxiliary decision-making module was added, and Mahalanobis distance was used to represent spatial relationships among samples, further improving the recognition effect. |
topic |
fault detection out-of-distribution convolutional neural network sound feature state detection |
url |
https://www.mdpi.com/2075-1702/9/8/179 |
work_keys_str_mv |
AT pengcui subhealthidentificationofreciprocatingmachinerybasedonsoundfeatureandooddetection AT jinjiawang subhealthidentificationofreciprocatingmachinerybasedonsoundfeatureandooddetection AT xiaobangli subhealthidentificationofreciprocatingmachinerybasedonsoundfeatureandooddetection AT chunfengli subhealthidentificationofreciprocatingmachinerybasedonsoundfeatureandooddetection |
_version_ |
1721192020476690432 |