Bearing Fault Degradation Modeling Based on Multitime Windows Fusion Unsupervised Health Indicator

Solichin, Mochammad; Noh, Yoojeong; Kang, Young-Jin

doi:10.1109/jsen.2023.3294361

Cited by 7 publications

(4 citation statements)

References 43 publications

Supporting

Mentioning

Contrasting

Order By: Relevance

“…These methods learn the relationships between the features extracted from labeled data and their corresponding labels to construct prediction models for predicting the RUL of unlabeled data. On the other hand, unsupervised ML methods, such as K-means clustering [104] and principal component analysis (PCA) [105], extract underlying patterns, structures, and similarities from unlabeled samples. These methods help to understand the characteristics of the data and the relationships among variables, providing valuable information for RUL prediction.…”

Section: Ml-based Approachmentioning

confidence: 99%

“…Zhao et al [115] utilized the S-transform and Gaussian pyramid to construct multi-scale features and performed dimensionality reduction using unsupervised techniques such as PCA with linear discriminant analysis for RUL estimation using a multiple linear regression model. Mochammad et al [105] extracted high-amplitude features from multi-time window data through frequency analysis, fused the features usings PCA, and generated highquality data for detecting abnormal states and estimating RUL. As shown in 13, this paper proposed a method for RUL prediction of bearings based on deep convolutional neural networks (DCNN) [118].…”

Section: Unsupervised ML For Remaining Life Predictionmentioning

confidence: 99%

See 1 more Smart Citation

Advancements in bearing remaining useful life prediction methods: a comprehensive review

Song,

Lin,

Jin

et al. 2024

Meas. Sci. Technol.

View full text Add to dashboard Cite

This paper presents a comprehensive review of the state-of-the-art techniques for predicting the remaining useful life (RUL) of rolling bearings. Four key aspects of bearing RUL prediction are considered: data acquiring, construction of health indicators (HI), development of RUL prediction algorithms, and evaluation of prediction results. Additionally, publicly available datasets that can be used to validate bearing prediction algorithms are described. The existing RUL prediction algorithms are categorized into three types and have been comprehensively reviewed: physical-based, statistical-based, and data-driven. In particular, the progress made in data-driven prediction methods is summarized, and typical methods such as RNN, TCN, GCN, Transform, and TL-based methods are introduced in detail. Finally, the challenges faced by data-driven methods in RUL prediction for bearings are discussed.

show abstract

Section: Ml-based Approachmentioning

confidence: 99%

Section: Unsupervised ML For Remaining Life Predictionmentioning

confidence: 99%

Advancements in bearing remaining useful life prediction methods: a comprehensive review

Song,

Lin,

Jin

et al. 2024

Meas. Sci. Technol.

View full text Add to dashboard Cite

show abstract

“…In this case, the HI is built using Artificial Intelligence (AI) techniques including Machine Learning (ML) [9], Deep Learning (DL) and Transfer Learning (TL) [2], etc. As an example, Mochammad et al [10] proposed an unsupervised health indicator obtained through frequency analysis and Principal Component Analysis (PCA) feature fusion. While these techniques can extract the degradation information without the a priori degradation mechanism knowledge, their accuracy strongly depends on the quality and quantity of the available data.…”

Section: Introductionmentioning

confidence: 99%

Machine Learning Approach for LPRE Bearings Remaining Useful Life Estimation Based on Hidden Markov Models and Fatigue Modelling

Galli,

Weber,

Hoblos

et al. 2024

Machines

View full text Add to dashboard Cite

Ball bearings are one of the most critical components of rotating machines. They ensure shaft support and friction reduction, thus their malfunctioning directly affects the machine’s performance. As a consequence, it is necessary to monitor the health conditions of such a component to avoid major degradations which could permanently damage the entire machine. In this context, HMS (Health Monitoring Systems) and PHM (Prognosis and Health Monitoring) methodologies propose a wide range of algorithms for bearing diagnosis and prognosis. The present article proposes an end-to-end PHM approach for ball bearing RUL (Remaining Useful Life) estimation. The proposed methodology is composed of three main steps: HI (Health Indicator) construction, bearing diagnosis and RUL estimation. The HI is obtained by processing non-stationary vibration data with the MODWPT (Maximum Overlap Discrete Wavelet Packet Transform). After that, a degradation profile is defined and coupled with crack initiation and crack propagation fatigue models. Lastly, a MB-HMM (Hidden Markov Model) is trained to capture the bearing degradation dynamics. This latter model is used to estimate the current degradation state as well as the RUL. The obtained results show good RUL prediction capabilities. In particular, the fatigue models allowed a reduction of the ML (Machine Learning) model size, improving the algorithms training phase.

show abstract

“…Zhao et al proposed a life prediction method based on a digital twin-driven model [ 10 ]. Mochammad et al designed a bearing fault degradation model based on a multitime window fusion unsupervised health indicator [ 11 ]. Zhang et al proposed a model-based analysis method for bearing daults [ 12 ].…”

Section: Introductionmentioning

confidence: 99%

A Reinforcement Ensemble Learning Method for Rolling Bearing Fault Diagnosis under Variable Work Conditions

Li,

Zhang,

Wang

et al. 2024

Sensors

View full text Add to dashboard Cite

Ensuring the smooth operation of rolling bearings requires a precise fault diagnosis. Particularly, identifying fault types under varying working conditions holds significant importance in practical engineering. Thus, we propose a reinforcement ensemble method for diagnosing rolling bearing faults under varying working conditions. Firstly, a reinforcement model was designed to select the optimal base learner. Stratified random sampling was used to extract four datasets from raw training data. The reinforcement model was trained by these four datasets, respectively, and we obtained four optimal base learners. Then, a sparse ANN was designed as the ensemble model and the reinforcement learning model that can successfully identify the fault type under variable work conditions was constructed. Extensive experiments were conducted, and the results demonstrate the superiority of the proposed method over other intelligent approaches, with significant practical engineering benefits.

show abstract

Bearing Fault Degradation Modeling Based on Multitime Windows Fusion Unsupervised Health Indicator

Cited by 7 publications

References 43 publications

Advancements in bearing remaining useful life prediction methods: a comprehensive review

Advancements in bearing remaining useful life prediction methods: a comprehensive review

Machine Learning Approach for LPRE Bearings Remaining Useful Life Estimation Based on Hidden Markov Models and Fatigue Modelling

A Reinforcement Ensemble Learning Method for Rolling Bearing Fault Diagnosis under Variable Work Conditions

Contact Info

Product

Resources

About