文献类型：期刊文献

英文题名：A novel multiscale hybrid neural network for intelligent fine-grained fault diagnosis

作者：Li, Chuanjiang Li, Shaobo Yang, Lei Wei, Hongjing Zhang, Ansi Zhang, Yizong

第一作者：Li, Chuanjiang

通信作者：Li, SB[1];Li, SB[2]

机构：[1]Guizhou Univ, Sch Mech Engn, Guiyang 550025, Guizhou, Peoples R China;[2]Guizhou Univ, State Key Lab Publ Big Data, Guiyang 550025, Guizhou, Peoples R China;[3]Guizhou Inst Technol, Sch Mech Engn, Guiyang 550003, Guizhou, Peoples R China

第一机构：Guizhou Univ, Sch Mech Engn, Guiyang 550025, Guizhou, Peoples R China

通信机构：corresponding author), Guizhou Univ, Sch Mech Engn, Guiyang 550025, Guizhou, Peoples R China;corresponding author), Guizhou Univ, State Key Lab Publ Big Data, Guiyang 550025, Guizhou, Peoples R China.

年份：2023

卷号：18

期号：1

起止页码：444-462

外文期刊名：NETWORKS AND HETEROGENEOUS MEDIA

收录：;EI(收录号：20230413427342);Scopus(收录号：2-s2.0-85146635593);WOS:【SCI-EXPANDED(收录号:WOS:000922644200018)】；

基金：This work was supported in part by the Guizhou Province Higher Education Project [No. QJH KY [2020]005, QJH KY [2020]009], and in part by China Scholarship Council [No. 202106670003]. Thanks for the computing support of the State Key Laboratory of Public Big Data, Guizhou University.

语种：英文

外文关键词：Fine-grained fault diagnosis; multiscale hybrid network; various working conditions; ResCNN; LSTM

摘要：Various intelligent methods for condition monitoring and fault diagnosis of mechanical equipment have been developed over the past few years. However, most of the existing deep learning (DL)-based fault diagnosis models perform well only when applied to deal with limited types of general failures, and these models fail to accurately distinguish fine-grained faults under multiple working conditions. To address these challenges, we propose a novel multiscale hybrid model (MSHM), which takes the raw vibration signal as input and progressively learns representative features containing both spatial and temporal information to effectively classify fine-grained faults in an end-to-end way. To simulate fine-grained failure scenarios in practice, more than 100 classes of faults under different working conditions are constructed based on two benchmark datasets, and the experimental results demonstrate that our proposed MSHM has advantages over state-of-the-art methods in terms of accuracy in identifying fine-grained faults, generality in handling fault classes of different granularity, and learning ability with limited data.