Fault Diagnosis and Health Management of Power Machinery

A special issue of Machines (ISSN 2075-1702). This special issue belongs to the section "Machines Testing and Maintenance".

Deadline for manuscript submissions: closed (30 November 2022) | Viewed by 49319

Special Issue Editors

Department of Industrial Engineering, Tsinghua University, Beijing 100084, China
Interests: machinery condition monitoring; intelligent fault diagnosis and prognostics; deep learning
College of Intelligence and Computing, Tianjin University, Tianjin 300072, China
Interests: machine learning; interpretable AI; fault diagnosis; condition monitoring

E-Mail Website
Guest Editor
School of Mechanical Engineering, Xi’an Jiaotong University, Xi’an 710049, China
Interests: sparse representation; machine learning; deep learning; condition monitoring

E-Mail Website
Guest Editor
Department of Design, Manufacturing and Engineering Management, University of Strathclyde, Glasgow G1 1XH, UK
Interests: diagnosis; prognosis; machine learning; reliability and asset management; condition monitoring

Special Issue Information

Dear Colleagues,

Modern power-machinery systems are typically operated in harsh operating and environmental conditions. Unexpected failures of such systems have been frequently reported and have severe consequences for production, businesses, and society, which leads to higher operating and maintenance costs. Therefore, it is of significance to develop a proactive program by which to effectively reduce unexpected failures and further improve the effectiveness of power-machinery operation. The advances in real-time-sensor monitoring techniques bring tremendous opportunities to enhance the reliability and safety of power-machinery systems. Particularly, the diagnosis process assists in the identification/classification of machinery faults in terms of severity and type. The knowledge from diagnosis is also utilized to quantify the machinery’s health state and track the evolution of machinery performance degradation in support of its remaining useful life (RUL) prognosis.

This Special Issue aims to collect original ideas for the fault diagnosis and prognosis of power-machinery systems. The guest editors invite original contributions on the following topics, but authors are not limited to these:

  • Constructing health indicators;
  • Sensor-data fusion techniques;
  • Condition monitoring and intelligent fault diagnosis;
  • Data-driven, physics-based and hybrid prognostic strategies;
  • Machine learning in fault diagnosis and prognosis;
  • The integration of diagnostic and prognostic decisions in maintenance strategies;
  • Uncertainty quantification.

Dr. Te Han
Dr. Ruonan Liu
Dr. Zhibin Zhao
Dr. Pradeep Kundu
Guest Editors

Manuscript Submission Information

Manuscripts should be submitted online at www.mdpi.com by registering and logging in to this website. Once you are registered, click here to go to the submission form. Manuscripts can be submitted until the deadline. All submissions that pass pre-check are peer-reviewed. Accepted papers will be published continuously in the journal (as soon as accepted) and will be listed together on the special issue website. Research articles, review articles as well as short communications are invited. For planned papers, a title and short abstract (about 100 words) can be sent to the Editorial Office for announcement on this website.

Submitted manuscripts should not have been published previously, nor be under consideration for publication elsewhere (except conference proceedings papers). All manuscripts are thoroughly refereed through a single-blind peer-review process. A guide for authors and other relevant information for submission of manuscripts is available on the Instructions for Authors page. Machines is an international peer-reviewed open access monthly journal published by MDPI.

Please visit the Instructions for Authors page before submitting a manuscript. The Article Processing Charge (APC) for publication in this open access journal is 2400 CHF (Swiss Francs). Submitted papers should be well formatted and use good English. Authors may use MDPI's English editing service prior to publication or during author revisions.

Keywords

  • condition monitoring
  • diagnosis
  • prognosis
  • condition-based maintenance
  • machine learning
  • power machinery

Benefits of Publishing in a Special Issue

  • Ease of navigation: Grouping papers by topic helps scholars navigate broad scope journals more efficiently.
  • Greater discoverability: Special Issues support the reach and impact of scientific research. Articles in Special Issues are more discoverable and cited more frequently.
  • Expansion of research network: Special Issues facilitate connections among authors, fostering scientific collaborations.
  • External promotion: Articles in Special Issues are often promoted through the journal's social media, increasing their visibility.
  • e-Book format: Special Issues with more than 10 articles can be published as dedicated e-books, ensuring wide and rapid dissemination.

Further information on MDPI's Special Issue polices can be found here.

Published Papers (18 papers)

Order results
Result details
Select all
Export citation of selected articles as:

Editorial

Jump to: Research

4 pages, 186 KiB  
Editorial
Fault Diagnosis and Health Management of Power Machinery
by Te Han, Ruonan Liu, Zhibin Zhao and Pradeep Kundu
Machines 2023, 11(4), 424; https://doi.org/10.3390/machines11040424 - 27 Mar 2023
Cited by 4 | Viewed by 1574
Abstract
Power-machinery systems are widely used in various industries, including manu-facturing, energy production, transportation, and infrastructure [...] Full article
(This article belongs to the Special Issue Fault Diagnosis and Health Management of Power Machinery)

Research

Jump to: Editorial

23 pages, 3974 KiB  
Article
Diesel Engine Fault Prediction Using Artificial Intelligence Regression Methods
by Denys P. Viana, Dionísio H. C. de Sá Só Martins, Amaro A. de Lima, Fabrício Silva, Milena F. Pinto, Ricardo H. R. Gutiérrez, Ulisses A. Monteiro, Luiz A. Vaz, Thiago Prego, Fabio A. A. Andrade, Luís Tarrataca and Diego B. Haddad
Machines 2023, 11(5), 530; https://doi.org/10.3390/machines11050530 - 5 May 2023
Cited by 4 | Viewed by 3623
Abstract
Predictive maintenance has been employed to reduce maintenance costs and production losses and to prevent any failure before it occurs. The framework proposed in this work performs diesel engine prognosis by evaluating the absolute value of the failure severity using random forest (RF) [...] Read more.
Predictive maintenance has been employed to reduce maintenance costs and production losses and to prevent any failure before it occurs. The framework proposed in this work performs diesel engine prognosis by evaluating the absolute value of the failure severity using random forest (RF) and multilayer perceptron (MLP) neural networks. A database was implemented with 3500 failure scenarios to overcome the problem of inducing destructive failures in diesel engines. Diesel engine failure signals were developed with the zero-dimensional thermodynamic model inside a cylinder coupled with the crankshaft torsional vibration model. Artificial neural networks and random forest regression models were employed for classifying and quantifying failures. The methodology was applied alongside an engine simulator to assess effectiveness and accuracy. The best-fitting performance was obtained with the random forest regressor with an RMSE value of 0.10 ± 0.03%. Full article
(This article belongs to the Special Issue Fault Diagnosis and Health Management of Power Machinery)
Show Figures

Figure 1

18 pages, 7676 KiB  
Article
Frequency Response Analysis for Three-Phase Star and Delta Induction Motors: Pattern Recognition and Fault Analysis Using Statistical Indicators
by Salem Mgammal Al-Ameri, Zulkurnain Abdul-Malek, Ali Ahmed Salem, Zulkarnain Ahmad Noorden, Ahmed Allawy Alawady, Mohd Fairouz Mohd Yousof, Mohamed Ibrahim Mosaad, Ahmed Abu-Siada and Hammam Abdurabu Thabit
Machines 2023, 11(1), 106; https://doi.org/10.3390/machines11010106 - 13 Jan 2023
Cited by 12 | Viewed by 2577
Abstract
This paper presents a new investigation to detect various faults within the three-phase star and delta induction motors (IMs) using a frequency response analysis (FRA). In this regard, experimental measurements using FRA are performed on three IMs of ratings 1 HP, 3 HP [...] Read more.
This paper presents a new investigation to detect various faults within the three-phase star and delta induction motors (IMs) using a frequency response analysis (FRA). In this regard, experimental measurements using FRA are performed on three IMs of ratings 1 HP, 3 HP and 5.5 HP in normal conditions, short-circuit fault (SC) and open-circuit fault (OC) conditions. The SC and OC faults are applied artificially between the turns (Turn-to-Turn), between the coils (Coil-to-Coil) and between the phases (Phase-to-Phase). The obtained measurements show that the star and delta IMs result in dissimilar FRA signatures for the normal and faulty windings. Various statistical indicators are used to quantify the deviations between the normal and faulty FRA signatures. The calculation is performed in three frequency ranges: low, middle and high ones, as the winding parameters including resistive, inductive and capacitive components dominate the frequency characteristics at different frequency ranges. Consequently, it is proposed that the boundaries for the used indicators facilitate fault identification and quantification. Full article
(This article belongs to the Special Issue Fault Diagnosis and Health Management of Power Machinery)
Show Figures

Figure 1

23 pages, 7514 KiB  
Article
Fourier-Based Adaptive Signal Decomposition Method Applied to Fault Detection in Induction Motors
by J. Jesus De Santiago-Perez, Martin Valtierra-Rodriguez, Juan Pablo Amezquita-Sanchez, Gerardo Israel Perez-Soto, Miguel Trejo-Hernandez and Jesus Rooney Rivera-Guillen
Machines 2022, 10(9), 757; https://doi.org/10.3390/machines10090757 - 1 Sep 2022
Cited by 3 | Viewed by 2508
Abstract
Time-frequency analysis is commonly used for fault detection in induction motors. A variety of signal decomposition techniques have been proposed in the literature, such as Wavelet transform, Empirical Mode Decomposition (EMD), Multiple Signal Classification (MUSIC), among others. They have been successfully used in [...] Read more.
Time-frequency analysis is commonly used for fault detection in induction motors. A variety of signal decomposition techniques have been proposed in the literature, such as Wavelet transform, Empirical Mode Decomposition (EMD), Multiple Signal Classification (MUSIC), among others. They have been successfully used in many works related with the topic. Nevertheless, the studied signals present amplitude changes and chirp-type frequency components that are difficult to track and isolate with the aforementioned techniques. The contribution of this work is the introduction of a novel technique for time-frequency signal decomposition that is based on an adaptive band-pass filter and the Short Time Fourier Transform (STFT), namely Fourier-Based Adaptive Signal Decomposition (FBASD) method. This method is capable of tracking and extracting nonstationary time-frequency components within a user-selected frequency band. With these components, a methodology for detecting and classifying broken rotor bars in induction motors using the startup transient current is also proposed. Full article
(This article belongs to the Special Issue Fault Diagnosis and Health Management of Power Machinery)
Show Figures

Figure 1

14 pages, 10154 KiB  
Article
Self-Attention and Multi-Task Based Model for Remaining Useful Life Prediction with Missing Values
by Kai Zhang and Ruonan Liu
Machines 2022, 10(9), 725; https://doi.org/10.3390/machines10090725 - 25 Aug 2022
Cited by 2 | Viewed by 1878
Abstract
Remaining useful life (RUL) prediction is recently a hot spot in industrial big data analysis research. It aims at obtaining the health status of the equipment in advance and making intelligent maintenance decisions. However, values missing is a common problem in real industrial [...] Read more.
Remaining useful life (RUL) prediction is recently a hot spot in industrial big data analysis research. It aims at obtaining the health status of the equipment in advance and making intelligent maintenance decisions. However, values missing is a common problem in real industrial applications which severely restricts the performance and application scope of RUL prediction. To deal with this problem, a novel prediction model called self-attention-based multi-task network (SMTN)is proposed. The spatiotemporal feature fusion module utilizes the self-attention mechanism and long short-term memory to fully exploit the information in space and time dimensions, multi-task learning module tries to learn a complete representation from incomplete data by performing the missing values imputation task, and the representation is simultaneously used for RUL prediction. Comparison experiments conducted on the C-MAPSS dataset verified the effectiveness of the proposed SMTN. Full article
(This article belongs to the Special Issue Fault Diagnosis and Health Management of Power Machinery)
Show Figures

Figure 1

14 pages, 2619 KiB  
Article
Power Transformer Condition Monitoring Based on Evaluating Oil Properties
by Ivan V. Bryakin, Igor V. Bochkarev, Vadim R. Khramshin, Vadim R. Gasiyarov and Ivan V. Liubimov
Machines 2022, 10(8), 630; https://doi.org/10.3390/machines10080630 - 29 Jul 2022
Cited by 5 | Viewed by 2757
Abstract
The authors review the techniques applied to diagnose oil aging. Further, the authors put forward a new diagnostic method. It stipulates for an additional high-frequency measuring loop formed in an operating transformer. This dielcometric measuring capacitor cell includes a stray capacitance generated by [...] Read more.
The authors review the techniques applied to diagnose oil aging. Further, the authors put forward a new diagnostic method. It stipulates for an additional high-frequency measuring loop formed in an operating transformer. This dielcometric measuring capacitor cell includes a stray capacitance generated by the transformer winding and core. The monitoring of the dependence between the physico–technological oil parameters and the measuring cell capacity is fundamental for the procedures for determining the composition and properties of the transformer oil filling this cell. High-frequency low-voltage is the output signal. To prevent the cross-impact of low-frequency high-voltage and high-frequency low-voltage circuits, the pilot high-frequency low-voltage is excited by a special coupling capacitor; the output to the power feeder is conducted through an appropriate low-frequency choke, where the measuring capacitor cell does not disturb the normal transformer operation. The key physical processes used for the monitoring are analyzed and described in detail. The authors develop an algorithm to compute the current resistances of both the transformer oil and its impurities. The transformer state is estimated by comparing the parameters specified with preset permissible limits. A structure flowchart based on two synchronous quadrature detectors is proposed for a high-frequency measuring loop. The monitoring system considered allows for determining the following insulating oil properties by using the algorithm for processing the recorded data: moisture content; dielectric losses due to the accumulation of aging products in the oil and its pollution; and the content of dissolved gases in the oil. The monitoring system operability and efficiency are confirmed by appropriate experimental studies. The experiments are conducted using a TM-25-6/0.4 oil-filled transformer with a capacity of 25 kVA in a steady-state operating mode at a load current of 25 A. It is found that the proposed control system allows for identifying a critical defect of increased moisture content in the oil with no more than 10% error, and a sensitivity threshold in the order of tenths of ppm. Full article
(This article belongs to the Special Issue Fault Diagnosis and Health Management of Power Machinery)
Show Figures

Figure 1

17 pages, 8100 KiB  
Article
A Quantile Dependency Model for Predicting Optimal Centrifugal Pump Operating Strategies
by Bruce Stephen, Blair Brown, Andrew Young, Andrew Duncan, Henrique Helfer-Hoeltgebaum, Graeme West, Craig Michie and Stephen D. J. McArthur
Machines 2022, 10(7), 557; https://doi.org/10.3390/machines10070557 - 10 Jul 2022
Cited by 1 | Viewed by 1593
Abstract
Used in many industrial applications, centrifugal pumps have optimal operating criteria specified at design. These criteria may not be precisely adhered to during operation which will ultimately reduce the life of the asset. Operators would therefore benefit from anticipating how often the design [...] Read more.
Used in many industrial applications, centrifugal pumps have optimal operating criteria specified at design. These criteria may not be precisely adhered to during operation which will ultimately reduce the life of the asset. Operators would therefore benefit from anticipating how often the design point is deviated from and hence how much asset degradation results. For centrifugal pumps, a novel set of covariates were proposed in this paper which formally partition observed operating zones with an Empirical Bivariate Quantile Partitioned distribution. This captured the dependency relation between operating parameters across plant configurations to predict the component wear that results from particular settings. The effectiveness of this was demonstrated through an operational case study in civil nuclear generation feedwater pumps where corroboration with bearing movements provides an indicator of plant wear. Such a technique is envisaged to inform operators of optimal plant configuration from multiple possibilities in advance of undertaking them. Full article
(This article belongs to the Special Issue Fault Diagnosis and Health Management of Power Machinery)
Show Figures

Figure 1

14 pages, 1396 KiB  
Communication
Semi-Supervised Transfer Learning Method for Bearing Fault Diagnosis with Imbalanced Data
by Xia Zong, Rui Yang, Hongshu Wang, Minghao Du, Pengfei You, Su Wang and Hao Su
Machines 2022, 10(7), 515; https://doi.org/10.3390/machines10070515 - 25 Jun 2022
Cited by 17 | Viewed by 2892
Abstract
Fault diagnosis is essential for assuring the safety and dependability of rotating machinery systems. Several emerging techniques, especially artificial intelligence-based technologies, are used to overcome the difficulties in this field. In most engineering scenarios, machines perform in normal conditions, which implies that fault [...] Read more.
Fault diagnosis is essential for assuring the safety and dependability of rotating machinery systems. Several emerging techniques, especially artificial intelligence-based technologies, are used to overcome the difficulties in this field. In most engineering scenarios, machines perform in normal conditions, which implies that fault data may be hard to acquire and limited. Therefore, the data imbalance and the deficiency of labels are practical challenges in the fault diagnosis of machinery bearings. Among the mainstream methods, transfer learning-based fault diagnosis is highly effective, as it transfers the results of previous studies and integrates existing resources. The knowledge from the source domain is transferred via Domain Adversarial Training of Neural Networks (DANN) while the dataset of the target domain is partially labeled. A semi-supervised framework based on uncertainty-aware pseudo-label selection (UPS) is adopted in parallel to improve the model performance by utilizing abundant unlabeled data. Through experiments on two bearing datasets, the accuracy of bearing fault classification surpassed the independent approaches. Full article
(This article belongs to the Special Issue Fault Diagnosis and Health Management of Power Machinery)
Show Figures

Figure 1

14 pages, 3706 KiB  
Article
Cross-Domain Remaining Useful Life Prediction Based on Adversarial Training
by Yuhang Duan, Jie Xiao, Honghui Li and Jie Zhang
Machines 2022, 10(6), 438; https://doi.org/10.3390/machines10060438 - 1 Jun 2022
Cited by 5 | Viewed by 2321
Abstract
Remaining useful life prediction can assess the time to failure of degradation systems. Currently, numerous neural network-based prediction methods have been proposed by researchers. However, most of the work contains an implicit prerequisite: the network training and testing data have the same operating [...] Read more.
Remaining useful life prediction can assess the time to failure of degradation systems. Currently, numerous neural network-based prediction methods have been proposed by researchers. However, most of the work contains an implicit prerequisite: the network training and testing data have the same operating conditions. To solve this problem, an adversarial discriminative domain adaption prediction method based on adversarial training is proposed to improve the accuracy of cross-domain prediction under different working conditions. First, an LSTM feature extraction network is constructed to mine the source domain data and the target domain data for deep feature representation. Subsequently, the parameters of the target domain feature extraction network are adjusted based on the idea of adversarial training to achieve domain invariant feature mining. The proposed scheme is experimented on a publicly available dataset and achieves state-of-the-art prediction performance compared to recent unsupervised domain adaptation prediction methods. Full article
(This article belongs to the Special Issue Fault Diagnosis and Health Management of Power Machinery)
Show Figures

Figure 1

13 pages, 4431 KiB  
Article
Morphological Component Analysis-Based Hidden Markov Model for Few-Shot Reliability Assessment of Bearing
by Yi Feng, Weijun Li, Kai Zhang, Xianling Li, Wenfang Cai and Ruonan Liu
Machines 2022, 10(6), 435; https://doi.org/10.3390/machines10060435 - 1 Jun 2022
Cited by 4 | Viewed by 1789
Abstract
Reliability is of great significance in ensuring the safe operation of modern industry, which mainly relies on data analysis and life tests. However, as the life of mechanical systems becomes increasingly longer with the rapid development of the manufacturing industry, the collection of [...] Read more.
Reliability is of great significance in ensuring the safe operation of modern industry, which mainly relies on data analysis and life tests. However, as the life of mechanical systems becomes increasingly longer with the rapid development of the manufacturing industry, the collection of historical failure data becomes progressively more time-consuming. In this paper, a few-shot reliability assessment approach is proposed in order to overcome the dependence on historical data. Firstly, the vibration response of a bearing was illustrated. Then, based on a vibration response analysis, a morphological component analysis (MCA) method based on sparse representation theory was used to decompose vibration signals and extract impulse signals. After the impulse components’ reconstruction, their statistical indexes were utilized as the input observation vector of a Mixture of Gaussians Hidden Markov Model (MoG-HMM) for a reliability estimation. Finally, the experimental dataset of an aerospace bearing was analyzed via the proposed method. The comparison results illustrate the effectiveness of the proposed method of a few-shot reliability assessment. Full article
(This article belongs to the Special Issue Fault Diagnosis and Health Management of Power Machinery)
Show Figures

Figure 1

11 pages, 3956 KiB  
Communication
Electromechanical Characteristics Analysis under DSISC Fault in Synchronous Generators
by Yu-Ling He, Ming-Hao Qiu, Xing-Hua Yuan, Xian-Long He, Hai-Peng Wang, Meng-Ya Jiang, Chris Gerada and Shu-Ting Wan
Machines 2022, 10(6), 432; https://doi.org/10.3390/machines10060432 - 1 Jun 2022
Cited by 1 | Viewed by 1996
Abstract
This paper studies the electromechanical characteristics of synchronous generators under dynamic stator interturn short circuit (DSISC). First, the air gap magnetic flux density (MFD) of the generator under normal and DSISC fault was obtained. Then, the expression for the phase current and the [...] Read more.
This paper studies the electromechanical characteristics of synchronous generators under dynamic stator interturn short circuit (DSISC). First, the air gap magnetic flux density (MFD) of the generator under normal and DSISC fault was obtained. Then, the expression for the phase current and the electromagnetic torque (EMT) were obtained. After this, the phase current and EMT were analyzed by finite element analysis (FEA). Finally, the measured electromechanical characteristics of the CS-5 generator under different conditions were analyzed in accordance with theory and simulation. It was shown that with the occurrence, and deterioration, of DSISC, the amplitude of the first harmonic, third harmonic and fifth harmonic of the phase current became more affected by the pulse. Meanwhile, the even-numbered harmonics components of EMT increased. Full article
(This article belongs to the Special Issue Fault Diagnosis and Health Management of Power Machinery)
Show Figures

Figure 1

24 pages, 14331 KiB  
Article
The Effect of Fit Clearance between Outer Race and Housing on Vibration Characteristics of a Cylindrical Roller Bearing with Localized Defects
by Fengtao Wang, Xin Ling, Zhen Zhang, Peng Dai, Shuping Yan and Lei Wang
Machines 2022, 10(6), 415; https://doi.org/10.3390/machines10060415 - 25 May 2022
Cited by 7 | Viewed by 2693
Abstract
Due to bolt looseness or operating ambient temperature, fit clearance can often be found between the outer ring and housing. The vibration characteristics of a cylindrical roller bearing with localized defects are greatly affected by the fit clearance and the accuracy of bearing [...] Read more.
Due to bolt looseness or operating ambient temperature, fit clearance can often be found between the outer ring and housing. The vibration characteristics of a cylindrical roller bearing with localized defects are greatly affected by the fit clearance and the accuracy of bearing fault diagnosis may be reduced. Thus, a mathematical model for a cylindrical roller bearing was constructed and the interaction between the outer ring and housing was described. The classical localized defects were modeled, such as the inner ring defect, outer ring defect and roller defect. The relative experiments were conducted to check the constructed model. Then, it was found that the RMS (Root Mean Square) of housing acceleration decreased with increasing housing stiffness and viscous damping. When the fit clearance and friction coefficient increased, the RMS values increased. Except for housing stiffness and viscous damping, there were no uniform change rules of defect frequency amplitudes for other conditions. In the bearing with an outer ring defect or roller defect, the shock times of housing acceleration and the contact force between outer ring and housing were delayed, while fit clearance decreased. However, contrary variation trends were found for the inner ring defect. If the phase difference between defect location and rotor unbalanced force increased, the RMS and acceleration fluctuation amplitudes for the inner ring defect decreased. When the location angle of the outer ring defect increased, RMS and frequency amplitudes increased and RMSy/RMS decreased. The calculated results may provide the theoretical foundation for condition monitoring rotating machinery systems. Full article
(This article belongs to the Special Issue Fault Diagnosis and Health Management of Power Machinery)
Show Figures

Figure 1

15 pages, 11861 KiB  
Article
A Two-Stage Transfer Regression Convolutional Neural Network for Bearing Remaining Useful Life Prediction
by Xianling Li, Kai Zhang, Weijun Li, Yi Feng and Ruonan Liu
Machines 2022, 10(5), 369; https://doi.org/10.3390/machines10050369 - 12 May 2022
Cited by 9 | Viewed by 2652
Abstract
Recently, deep learning techniques have been successfully used for bearing remaining useful life (RUL) prediction. However, the degradation pattern of bearings can be much different from each other, which leads to the trained model usually not being able to work well for RUL [...] Read more.
Recently, deep learning techniques have been successfully used for bearing remaining useful life (RUL) prediction. However, the degradation pattern of bearings can be much different from each other, which leads to the trained model usually not being able to work well for RUL prediction of a new bearing. As a method that can adapt a model trained on source datasets to a different but relative unlabeled target dataset, transfer learning shows the potential to solve this problem. Therefore, we propose a two-stage transfer regression (TR)-based bearing RUL prediction method. Firstly, the incipient fault point (IFP) is detected by a convolutional neural network (CNN) classifier to identity the start time of degradation stage and label the training samples. Then, a transfer regression CNN with multiloss is constructed for RUL prediction, including regression loss, classification loss, maximum mean discrepancy (MMD) and regularization loss, which can not only extract fault information from fault classification loss for RUL prediction, but also minimize the probability distribution distance, thus helping the method to be trained in a domain-invariant way via the transfer regression algorithm. Finally, real data collected from run-to-failure bearing experiments are analyzed by the TR-based CNN method. The results and comparisons with state-of-the-art methods demonstrate the superiority and reliable performance of the proposed method for bearing RUL prediction. Full article
(This article belongs to the Special Issue Fault Diagnosis and Health Management of Power Machinery)
Show Figures

Figure 1

17 pages, 10599 KiB  
Article
Bearing Fault Diagnosis via Incremental Learning Based on the Repeated Replay Using Memory Indexing (R-REMIND) Method
by Junhui Zheng, Hui Xiong, Yuchang Zhang, Kaige Su and Zheyuan Hu
Machines 2022, 10(5), 338; https://doi.org/10.3390/machines10050338 - 6 May 2022
Cited by 8 | Viewed by 2218
Abstract
In recent years, deep-learning schemes have been widely and successfully used to diagnose bearing faults. However, as operating conditions change, the distribution of new data may differ from that of previously learned data. Training using only old data cannot guarantee good performance when [...] Read more.
In recent years, deep-learning schemes have been widely and successfully used to diagnose bearing faults. However, as operating conditions change, the distribution of new data may differ from that of previously learned data. Training using only old data cannot guarantee good performance when handling new data, and vice versa. Here, we present an incremental learning scheme based on the Repeated Replay using Memory Indexing (R-REMIND) method for bearing fault diagnosis. R-REMIND can learn new information under various working conditions while retaining older information. First, we use a feature extraction network similar to the Inception-v4 neural network to collect bearing vibration data. Second, we encode the features by product quantization and store the features in indices. Finally, the parameters of the feature extraction and classification networks are updated using real and reconstructed features, and the model did not forget old information. The experiment results show that the R-REMIND model exhibits continuous learning ability with no catastrophic forgetting during sequential tasks. Full article
(This article belongs to the Special Issue Fault Diagnosis and Health Management of Power Machinery)
Show Figures

Figure 1

18 pages, 6570 KiB  
Article
Imbalanced Fault Diagnosis of Rolling Bearing Using Data Synthesis Based on Multi-Resolution Fusion Generative Adversarial Networks
by Chuanzhu Hao, Junrong Du and Haoran Liang
Machines 2022, 10(5), 295; https://doi.org/10.3390/machines10050295 - 22 Apr 2022
Cited by 11 | Viewed by 2387
Abstract
Fault diagnosis of industrial bearings plays an invaluable role in the health monitoring of rotating machinery. In practice, there is far more normal data than faulty data, so the data usually exhibit a highly skewed class distribution. Algorithms developed using unbalanced datasets will [...] Read more.
Fault diagnosis of industrial bearings plays an invaluable role in the health monitoring of rotating machinery. In practice, there is far more normal data than faulty data, so the data usually exhibit a highly skewed class distribution. Algorithms developed using unbalanced datasets will suffer from severe model bias, reducing the accuracy and stability of the classification algorithm. To address these issues, a novel Multi-resolution Fusion Generative Adversarial Network (MFGAN) is proposed for the imbalanced fault diagnosis of rolling bearings via data augmentation. In the data-generation process, the improved feature transfer-based generator receives normal data as input to better learn the fault features, mapping the normal data into fault data space instead of random data space. A multi-scale ensemble discriminator architecture is designed to replace original single discriminator structure in the discriminative process, and multi-scale features are learned via ensemble discriminators. Finally, the proposed framework is validated on the public bearing dataset from Case Western Reserve University (CWRU), and experimental results show the superiority of our method. Full article
(This article belongs to the Special Issue Fault Diagnosis and Health Management of Power Machinery)
Show Figures

Figure 1

15 pages, 6651 KiB  
Article
Vibration Responses of the Bearing-Rotor-Gear System with the Misaligned Rotor
by Fengtao Wang, Peng Dai, Jianping Wang and Linkai Niu
Machines 2022, 10(4), 267; https://doi.org/10.3390/machines10040267 - 8 Apr 2022
Cited by 7 | Viewed by 3161
Abstract
The bearing-rotor-gear system is an important mechanical component for transmitting motion and power. Due to the complex responses, the condition assessment of the transmission system becomes more difficult. Thus, a model of the bearing-rotor-gear system with a misaligned rotor is built for implementing [...] Read more.
The bearing-rotor-gear system is an important mechanical component for transmitting motion and power. Due to the complex responses, the condition assessment of the transmission system becomes more difficult. Thus, a model of the bearing-rotor-gear system with a misaligned rotor is built for implementing the complex response analysis. The misalignment rotor is realized by offset connection of couplings, and the creative excitation force is transferred to the bearing inner ring through the rotor. The constructed model is checked by the corresponding experiment. From the simulation results, it is found that vibration responses are modulated by rotor frequencies, and there are rotor frequencies, harmonic frequencies of bearings, and gear pairs in the spectrum. When the misalignment defect is deepening, the high-order harmonic responses are excited. If the revolving speed increases, the modulation of the rotor frequencies is accentuated, the vibration intensity generated by gear pairs is attenuated, while the harmonic response and super-harmonic response of bearings can be suppressed, and the system vibrates mainly at the low-frequency band. When the load becomes higher, the amplitudes of the rotor frequencies, meshing frequencies, and defect frequencies are all increased. Full article
(This article belongs to the Special Issue Fault Diagnosis and Health Management of Power Machinery)
Show Figures

Figure 1

20 pages, 9649 KiB  
Article
A Novel Impact Feature Extraction Method Based on EMD and Sparse Decomposition for Gear Local Fault Diagnosis
by Zhongze Liu, Kang Ding, Huibin Lin, Guolin He, Canyi Du and Zhuyun Chen
Machines 2022, 10(4), 242; https://doi.org/10.3390/machines10040242 - 30 Mar 2022
Cited by 31 | Viewed by 3491
Abstract
Sparse decomposition has been widely used in gear local fault diagnosis due to its outstanding performance in feature extraction. The extraction results depend heavily on the similarity between dictionary atoms and fault feature signal. However, the transient impact signal aroused by gear local [...] Read more.
Sparse decomposition has been widely used in gear local fault diagnosis due to its outstanding performance in feature extraction. The extraction results depend heavily on the similarity between dictionary atoms and fault feature signal. However, the transient impact signal aroused by gear local defect is usually submerged in meshing harmonics and noise. It is still a challenging task to construct high-quality impact dictionary for complex actual signal. To handle this issue, a novel impact feature extraction method based on Empirical Mode Decomposition (EMD) and sparse decomposition is proposed in this paper. Firstly, EMD is employed to adaptively decompose the original signal into several Intrinsic Mode Functions (IMFs). The high-frequency resonance component is separated from meshing harmonics and part of the noise. Then, the IMF with the prominent impact features is selected as the Main Intrinsic Mode Function (MIMF) based on the kurtosis. Accordingly, the modal parameters required for impact dictionary are identified from the MIMF by correlation filtering. Finally, the transient impact component is extracted from the original signal by Match Pursuit (MP). The proposed method was adequately evaluated by a gear local fault simulation signal, and the single-stage gearbox and five-speed transmission experiments. The effectiveness and superiority of the proposed method is validated by comparison with other feature extraction techniques. Full article
(This article belongs to the Special Issue Fault Diagnosis and Health Management of Power Machinery)
Show Figures

Figure 1

23 pages, 68117 KiB  
Article
A p−V Diagram Based Fault Identification for Compressor Valve by Means of Linear Discrimination Analysis
by Xueying Li, Peng Ren, Zhe Zhang, Xiaohan Jia and Xueyuan Peng
Machines 2022, 10(1), 53; https://doi.org/10.3390/machines10010053 - 10 Jan 2022
Cited by 12 | Viewed by 3802
Abstract
The pressure-volume diagram (p−V diagram) is an established method for analyzing the thermodynamic process in the cylinder of a reciprocating compressor as well as the fault of its core components including valves. The failure of suction/discharge valves is the most common cause of [...] Read more.
The pressure-volume diagram (p−V diagram) is an established method for analyzing the thermodynamic process in the cylinder of a reciprocating compressor as well as the fault of its core components including valves. The failure of suction/discharge valves is the most common cause of unscheduled shutdowns, and undetected failure may lead to catastrophic accidents. Although researchers have investigated fault classification by various estimation techniques and case studies, few have looked deeper into the barriers and pathways to realize the level determination of faults. The initial stage of valve failure is characterized in the form of mild leakage; if this is identified at this period, more serious accidents can be prevented. This study proposes a fault diagnosis and severity estimation method of the reciprocating compressor valve by virtue of features extracted from the p−V diagram. Four-dimensional characteristic variables consisting of the pressure ratio, process angle coefficient, area coefficient, and process index coefficient are extracted from the p−V diagram. Principal Component Analysis (PCA) and Linear Discriminant Analysis (LDA) were applied to establish the diagnostic model, where PCA realizes feature amplification and projection, then LDA implements feature dimensionality reduction and failure prediction. The method was validated by the diagnosis of various levels of severity of valve leakage in a reciprocating compressor, and further, applied in the diagnosis of two actual faults: Mild leakage caused by the cracked valve plate in a reciprocating compressor, and serious leakage caused by the deformed valve in a hydraulically driven piston compressor for a hydrogen refueling station (HRS). Full article
(This article belongs to the Special Issue Fault Diagnosis and Health Management of Power Machinery)
Show Figures

Graphical abstract

Back to TopTop