An information-based K-singular-value decomposition method for rolling element bearing diagnosis

Liang, Kaixuan; Zhao, Ming; Lin, Jing; Jiao, Jinyang

doi:10.1016/j.isatra.2019.06.012

Cited by 25 publications

(12 citation statements)

References 35 publications

Supporting

Mentioning

Contrasting

Order By: Relevance

“…Therefore, fault detection is a widely explored and emerging research area in machine condition monitoring. The current fault detection methods can be divided into four categories according to their development: methods based on (1) signal processing, (2) physical models, (3) machine learning, and (4) hybrid approach (Haidong et al., 2020; Liang et al., 2020). The data‐driven fault detection methods such as machine learning deliver remarkable performance.…”

Section: Related Workmentioning

confidence: 99%

Equipment activity recognition and early fault detection in automated construction through a hybrid machine learning framework

Harichandran

Raphael

Mukherjee

2022

Computer aided Civil Eng

View full text Add to dashboard Cite

Existing studies on automated construction equipment monitoring have focused mainly on activity recognition rather than fault detection. This paper proposes a novel equipment activity recognition and fault detection framework called hybrid unsupervised and supervised machine learning (HUS‐ML). HUS‐ML first identifies normal operations and known faulty conditions through supervised learning. Then, an anomaly detection algorithm is applied to spot any unseen faulty conditions. The framework is tested using acceleration measurements from a low‐rise automated construction system prototype. HUS‐ML outperformed the conventional machine learning approach in activity recognition and fault detection with an average F1 score of 86.6%. The conventional approach failed to detect unseen faulty operations. HUS‐ML identified known faulty operations and unseen faulty operations with F1 scores of 98.11% and 76.19%, respectively. The generalizability of the framework is demonstrated by validating it on an independent benchmark dataset with good results.

show abstract

Section: Related Workmentioning

confidence: 99%

Equipment activity recognition and early fault detection in automated construction through a hybrid machine learning framework

Harichandran

Raphael

Mukherjee

2022

Computer aided Civil Eng

View full text Add to dashboard Cite

show abstract

“…Based on the vibration analysis, various signal processing techniques have been developed for the feature extraction and fault detection of rotating machines, such as empirical mode decomposition (EMD) [8][9], variational mode decomposition (VMD) [10][11], wavelet transform (WT) [12], orthogonal matching pursuit (OMP) [13], singular value decomposition (SVD) [14][15], machine learning (ML) [16][17], etc. EMD aims to adaptively process the nonstationary signal by decomposing the signal into a series of intrinsic mode functions (IMFs).…”

Section: Cms-tkeomentioning

confidence: 99%

Improved cyclostationary analysis method based on TKEO and its application on the faults diagnosis of induction motors

Wang

Yang

et al. 2022

ISA Transactions

View full text Add to dashboard Cite

“…To assure the reliable and stable operation of mechanical equipment, detecting faults in time is very significant [1]. Owing to the vibration signals of rolling bearings containing abundant fault information, vibration-based methods have been a basic approach in extracting the bearing fault feature [2,3]. When the roller regularly passing through the local fault in the bearing, the recurrent force shocks excite high frequency resonance and present in the vibration signal [4].…”

Section: Introductionmentioning

confidence: 99%

Multi-period synchronous deconvolution for complex fault diagnosis of rolling element bearings

Zhou

Zhao

et al. 2023

Preprint

View full text Add to dashboard Cite

Due to inaccurate period knowledge or incomplete information, there are limitations in extracting all potential complex faults of rolling bearings and enhancing weak faults. To address this problem, this article introduces a novel deconvolution method, named multi-period synchronous deconvolution (MPSD). In this method, decomposing matrix-matrix in solving the filter coefficients are used to replace designing objective function to avoid only obtaining the filter signal of the optimal objective function. Based on eliminating the effects of subspace noise, a novel evaluation index, called characteristic information ratio, is proposed to evaluate fault significance by fault information levels instead of energy. In addition, an informative subspace selection strategy is proposed to control the weighting coefficients of each subspace in reconstructed signal. Without the predetermined fault periods, the proposed MPSD can simultaneously extract latent multiple faults and enhance weak fault features. Finally, simulations and experimental cases substantiate the efficacy and eminence of MPSD.

show abstract

An information-based K-singular-value decomposition method for rolling element bearing diagnosis

Cited by 25 publications

References 35 publications

Equipment activity recognition and early fault detection in automated construction through a hybrid machine learning framework

Equipment activity recognition and early fault detection in automated construction through a hybrid machine learning framework

Improved cyclostationary analysis method based on TKEO and its application on the faults diagnosis of induction motors

Multi-period synchronous deconvolution for complex fault diagnosis of rolling element bearings

Contact Info

Product

Resources

About