Fast time-frequency manifold learning and its reconstruction for transient feature extraction in rotating machinery fault diagnosis

Ding, Xiaoxi; Li, Quanchang; Lin, Lun; He, Qingbo; Shao, Yimin

doi:10.1016/j.measurement.2019.04.030

Cited by 58 publications

(20 citation statements)

References 29 publications

Supporting

Mentioning

Contrasting

Order By: Relevance

“…The PCA is an unsupervised linear method that allows reducing the dimension of a correlated data set to a linear space of unrelated indicators while preserving the greatest possible variance. The application of the PCA results in the extraction of a new set of features that are sorted according to the cumulative variance that preserves and are known as principal components [5,11,42]. On the other hand, the LDA is a supervised linear method that allows a dimensionality reduction by extracting a new set of features, in which the maximization of the data separability is achieved for a C number of considered classes.…”

Section: Machine Learning-based Feature Reductionmentioning

confidence: 99%

Condition Monitoring Method for the Detection of Fault Graduality in Outer Race Bearing Based on Vibration-Current Fusion, Statistical Features and Neural Network

et al. 2021

View full text Add to dashboard Cite

Bearings are the elements that allow the rotatory movement in induction motors, and the fault occurrence in these elements is due to excessive working conditions. In induction motors, electrical erosion remains the most common phenomenon that damages bearings, leading to incipient faults that gradually increase to irreparable damages. Thus, condition monitoring strategies capable of assessing bearing fault severities are mandatory to overcome this critical issue. The contribution of this work lies in the proposal of a condition monitoring strategy that is focused on the analysis and identification of different fault severities of the outer race bearing fault in an induction motor. The proposed approach is supported by fusion information of different physical magnitudes and the use of Machine Learning and Artificial Intelligence. An important aspect of this proposal is the calculation of a hybrid-set of statistical features that are obtained to characterize vibration and stator current signals by its processing through domain analysis, i.e., time-domain and frequency-domain; also, the fusion of information of both signals by means of the Linear Discriminant Analysis is important due to the most discriminative and meaningful information is retained resulting in a high-performance condition characterization. Besides, a Neural Network-based classifier allows validating the effectiveness of fusion information from different physical magnitudes to face the diagnosis of multiple fault severities that appear in the bearing outer race. The method is validated under an experimental data set that includes information related to a healthy condition and five different severities that appear in the outer race of bearings.

show abstract

Section: Machine Learning-based Feature Reductionmentioning

confidence: 99%

Condition Monitoring Method for the Detection of Fault Graduality in Outer Race Bearing Based on Vibration-Current Fusion, Statistical Features and Neural Network

et al. 2021

View full text Add to dashboard Cite

show abstract

“…Different from STFT, short-frequency Fourier transform (SFFT) 27,28 is applied in this study to transform IMFs into sub-TFDs more effectively. Most importantly, because only a limited frequency band is transformed, the size of matrix analyzed in the subsequent manifold learning can be greatly reduced.…”

Section: Variational Mode Manifold Reinforcement Learningmentioning

confidence: 99%

Time-frequency synthesis analysis for complex signal of rotating machinery via variational mode manifold reinforcement learning

Ding

Wang³

et al. 2019

Proceedings of the Institution of Mechanical Engineers, Part C:

Self Cite

View full text Add to dashboard Cite

The transient signal caused by localized fault in rotating machinery always contains complex modulation information with heavy background noise distributed, which brings much difficulties for fault feature identification in the application of rotating machinery fault diagnosis. Focusing on the sensitive feature extraction from these complex signals, this paper proposes a novel variational mode manifold reinforcement learning (VM2RL) to adaptively construct time-frequency synthesis analysis for enhancement of transient features. First, a method of adaptive variational mode decomposition (VMD) is employed to divide the raw spectrum of the given signal into several sub-bands with different frequency modulation information. Second, an improved time-frequency manifold (ITFM) learning is introduced to gain the topological manifold structure from those sub-distributions in time-frequency domain. Then, a sound-enhanced signature of transient features on the whole time-frequency plane can be synthesized by combining those sub-TFMs from each modulated segment back to the corresponding frequency band. Finally, the time-frequency envelope spectrum for fault diagnosis is further obtained through statistically evaluating their amplitude distribution. Among them, short-frequency Fourier transform (SFFT) is introduced to transform local frequency bands into a series of TFDs which improves the computational efficiency of TFM learning. In this manner, the desired transient distribution on full time-frequency plane can be automatically reconstructed by VM2RL with manifold reinforced in a data-driven way. A simulation study and two experimental signals are both analyzed here, and fast spectral kurtosis and conventional VMD methods are also used to verify its effectiveness. Meanwhile, a quantitative analysis has been provided to further illustrate its superiority in the application of complex signal fault of rotating machinery.

show abstract

“…The principle of selection is to try not to miss a feature that may be useful, but not to abuse too many features. To extract the features, many signal processing methods have been used in the area of rotating machine health monitoring and diagnosis, such as time-domain and frequency-domain feature parameters processing [ 8 , 9 , 10 ], discrete wavelet transform (DWT) [ 11 ], empirical mode decomposition (EMD) [ 12 ], time-frequency analysis (TFA) [ 13 ], Mel-frequency cepstrum (MFC) [ 14 ], and Shannon entropy [ 15 ]. Among them, Shannon entropy features have been widely used in machine health monitoring recently.…”

Section: Introductionmentioning

confidence: 99%

A Novel End-To-End Feature Selection and Diagnosis Method for Rotating Machinery

Wang

Zhao

Zhang

et al. 2021

Sensors

View full text Add to dashboard Cite

Feature selection is to obtain effective features from data, also known as feature engineering. Traditional feature selection and predictive model learning are separated, and there is a problem of inconsistency of criteria. This paper presents an end-to-end feature selection and diagnosis method that organically unifies feature expression learning and machine prediction learning into one model. The algorithm first combines the prediction model to calculate the mean impact value (MIVs) of the feature and realizes primary feature selection for the prediction model by selecting the feature with a larger MIV. In order to take into account the performance of the feature itself, the within-class and between-class discriminant analysis (WBDA) method is proposed, and combined with the feature diversity strategy, the feature-oriented secondary selection is realized. Eventually, feature vectors obtained by two selections are classified using a multi-class support vector machine (SVM). Compared with the modified network variable selection algorithm (MIVs), the principal component analysis dimensionality reduction algorithm (PCA), variable selection based on compensative distance evaluation technology (CDET), and other algorithms, the proposed method MIVs-WBDA exhibits excellent classification accuracy owing to the fusion of feature selection and predictive model learning. According to the results of classification accuracy testing after dimensionality reduction on rotating machinery status, the MIVs-WBDA method has a 3% classification accuracy improvement under the low-dimensional feature set. The typical running time of this classification learning algorithm is less than 10 s, while using deep learning, its running time will be more than a few hours.

show abstract

Fast time-frequency manifold learning and its reconstruction for transient feature extraction in rotating machinery fault diagnosis

Cited by 58 publications

References 29 publications

Condition Monitoring Method for the Detection of Fault Graduality in Outer Race Bearing Based on Vibration-Current Fusion, Statistical Features and Neural Network

Condition Monitoring Method for the Detection of Fault Graduality in Outer Race Bearing Based on Vibration-Current Fusion, Statistical Features and Neural Network

Time-frequency synthesis analysis for complex signal of rotating machinery via variational mode manifold reinforcement learning

A Novel End-To-End Feature Selection and Diagnosis Method for Rotating Machinery

Contact Info

Product

Resources

About