Machine Fault Detection Using a Hybrid CNN-LSTM Attention-Based Model
Open accessThe DOI matches its Crossref record; 61 of the 61 references it cites were found in a registry or an index.
Abstract
The predictive maintenance of electrical machines is a critical issue for companies, as it can greatly reduce maintenance costs, increase efficiency, and minimize downtime. In this paper, the issue of predicting electrical machine failures by predicting possible anomalies in the data is addressed through time series analysis. The time series data are from a sensor attached to an electrical machine (motor) measuring vibration variations in three axes: X (axial), Y (radial), and Z (radial X). The dataset is used to train a hybrid convolutional neural network with long short-term memory (CNN-LSTM) architecture. By employing quantile regression at the network output, the proposed approach aims to manage the uncertainties present in the data. The application of the hybrid CNN-LSTM attention-based model, combined with the use of quantile regression to capture uncertainties, yielded superior results compared to traditional reference models. These results can benefit companies by optimizing their maintenance schedules and improving the overall performance of their electric machines.
Overview
The claims and statistics of this paper haven’t been checked yet.
Free: it reads the open-access text and checks the first 5 claims that cite a source we can check. Create a free account to check every claim.
Integrity signals
Registry
Cites
Statistics
Verification
Claims the paper makes
The claims and statistics of this paper haven’t been checked yet.
Free: it reads the open-access text and checks the first 5 claims that cite a source we can check. Create a free account to check every claim.
Self-check
Percentages, sums and reported differences re-computed from the paper’s own numbers.
The self-check needs the paper’s full text.
Citations
Cites
Temporal Fusion Transformers for interpretable multi-horizon time series forecasting
Lim, B., Arık, S. Ö., Loeff, N., & Pfister, T. (2021) · International Journal of Forecasting · doi:10.1016/j.ijforecast.2021.03.012 · Cited by 3,291
A Comprehensive Survey on Graph Anomaly Detection With Deep Learning
Ma, X., Wu, J., Xue, S., Yang, J., Zhou, C., Sheng, Q. Z., Xiong, H., & Akoglu, L. (2023) · IEEE Transactions on Knowledge and Data Engineering · doi:10.1109/tkde.2021.3118815 · Cited by 835 · cited 1×
“A major component of predictive maintenance is anomaly detection, which enables businesses to spot possible breakdowns quickly [9].”
BackgroundWang, Z., Liu, M., Yan, W.-J., Song, H., Zhou, Z., Yuen, K.-V., Wei, Q., & Cheng, S. S. (2021) · IEEE Transactions on Instrumentation and Measurement · doi:10.1109/tim.2021.3123218 · Cited by 678 · cited 1×
“The application of EWT for failure diagnosis extends to other types of machines, such as wind turbines [42], and other forecasting applications [43].”
BackgroundWavelet Transform Application for/in Non-Stationary Time-Series Analysis: A Review
Rhif, M., Ben Abbes, A., Farah, I. R., Martínez, B., & Sang, Y. (2019) · Applied Sciences · doi:10.3390/app9071345 · Cited by 598 · cited 1×
“The model has more effective results because it focuses on the variation trend and not on the signal noise [31].”
BackgroundA review on maintenance optimization
de Jonge, B., & Scarf, P. A. (2020) · European Journal of Operational Research · doi:10.1016/j.ejor.2019.09.047 · Cited by 567 · cited 1×
“Identifying anomalous behavior in equipment is increasingly recognized as a crucial factor in anticipating maintenance actions [3] and achieving gains by avoiding unplanned downtime [4].”
BackgroundListou Ellefsen, A., Bjørlykhaug, E., Æsøy, V., Ushakov, S., & Zhang, H. (2019) · Reliability Engineering & System Safety · doi:10.1016/j.ress.2018.11.027 · Cited by 523 · cited 1×
“Machine learning algorithms [18], such as decision trees [19], can be used to identify patterns in sensor readings and make predictions based on this information, while deep learning techniques have been widely used [20].”
BackgroundMachine learning advances for time series forecasting
Masini, R. P., Medeiros, M. C., & Mendes, E. F. (2023) · Journal of Economic Surveys · doi:10.1111/joes.12429 · Cited by 515
A new dynamic predictive maintenance framework using deep learning for failure prognostics
Nguyen, K. T. P., & Medjaher, K. (2019) · Reliability Engineering & System Safety · doi:10.1016/j.ress.2019.03.018 · Cited by 385 · cited 1×
“Identifying anomalous behavior in equipment is increasingly recognized as a crucial factor in anticipating maintenance actions [3] and achieving gains by avoiding unplanned downtime [4].”
BackgroundWell production forecasting based on ARIMA-LSTM model considering manual operations
Fan, D., Sun, H., Yao, J., Zhang, K., Yan, X., & Sun, Z. (2021) · Energy · doi:10.1016/j.energy.2020.119708 · Cited by 379 · cited 1×
“[51], combining the LSTM and ARIMA models.”
Wu, J., Hu, K., Cheng, Y., Zhu, H., Shao, X., & Wang, Y. (2020) · ISA Transactions · doi:10.1016/j.isatra.2019.07.004 · Cited by 349 · cited 1×
“Time series spectrum analysis can be used to identify changes in the machine’s electrical signals, which may indicate failures in internal components such as bearings or windings [21].”
BackgroundAli, M. Z., Shabbir, M. N. S. K., Liang, X., Zhang, Y., & Hu, T. (2019) · IEEE Transactions on Industry Applications · doi:10.1109/tia.2019.2895797 · Cited by 349 · cited 1×
“One is to use vibration analysis techniques to detect changes in the vibration signature of an engine [17], which can indicate misalignment, excessive wear, or other mechanical problems.”
MethodChen, Y., Cao, R., Chen, J., Liu, L., & Matsushita, B. (2021) · ISPRS Journal of Photogrammetry and Remote Sensing · doi:10.1016/j.isprsjprs.2021.08.015 · Cited by 303 · cited 1×
“The filter works by fitting a polynomial of a specified order to a local window of the data and using this polynomial to estimate the smoothed values at each point in the time series [69].”
BackgroundKedadouche, M., Thomas, M., & Tahan, A. (2016) · Mechanical Systems and Signal Processing · doi:10.1016/j.ymssp.2016.02.049 · Cited by 295 · cited 1×
“Its flexibility in selecting wavelet functions allows for the optimal representation of the signal under analysis, further increasing its effectiveness in a wide range of applications [67].”
BackgroundDeng, W., Zhang, S., Zhao, H., & Yang, X. (2018) · IEEE Access · doi:10.1109/access.2018.2834540 · Cited by 282
Fault diagnosis of induction motor based on decision trees and adaptive neuro-fuzzy inference
Tran, V. T., Yang, B.-S., Oh, M.-S., & Tan, A. C. C. (2009) · Expert Systems with Applications · doi:10.1016/j.eswa.2007.12.010 · Cited by 224
Stefenon, S. F., Seman, L. O., Mariani, V. C., & Coelho, L. D. S. (2023) · Energies · doi:10.3390/en16031371 · Cited by 148
Self-citationThirumala, K., Pal, S., Jain, T., & Umarikar, A. C. (2019) · Neurocomputing · doi:10.1016/j.neucom.2019.01.038 · Cited by 137 · cited 1×
“EWT enables the extraction of information from signals with low signal-to-noise ratios [66].”
BackgroundStefenon, S. F., Seman, L. O., Aquino, L. S., & Coelho, L. D. S. (2023) · Energy · doi:10.1016/j.energy.2023.127350 · Cited by 129 · cited 1×
“The output gate ot controls which part of the candidate memory cell ct is passed through the hyperbolic tangent activation function to produce the hidden state ht [70].”
BackgroundSelf-citationLong, Z., Zhang, X., Zhang, L., Qin, G., Huang, S., Song, D., Shao, H., & Wu, G. (2020) · Measurement · doi:10.1016/j.measurement.2020.108718 · Cited by 123
Yu, H., Li, H., & Li, Y. (2020) · ISA Transactions · doi:10.1016/j.isatra.2020.07.025 · Cited by 112 · cited 1×
“[62] presented a novel vibration signal fusion approach combining the improved EWT and the variance contribution rate.”
MethodLi, D., Tan, Y., Zhang, Y., Miao, S., & He, S. (2022) · International Journal of Electrical Power & Energy Systems · doi:10.1016/j.ijepes.2022.108743 · Cited by 95 · cited 1×
“[49], the TFT can improve the reliability and compactness of the forecasting and can even be applied to medium-term hourly time series data.”
BackgroundNeto, N. F. S., Stefenon, S. F., Meyer, L. H., Ovejero, R. G., & Leithardt, V. R. Q. (2022) · Sensors · doi:10.3390/s22166121 · Cited by 93 · cited 1×
“us to look at how a given statistic changes over time [10].”
BackgroundSelf-citationLite and Efficient Deep Learning Model for Bearing Fault Diagnosis Using the CWRU Dataset
Yoo, S.-B., Jo, J.-H., & Ban, S.-W. (2023) · Sensors · doi:10.3390/s23063157 · Cited by 91
Optimized EWT-Seq2Seq-LSTM with Attention Mechanism to Insulators Fault Prediction
Klaar, A. C. R., Stefenon, S. F., Seman, L. O., Mariani, V. C., & Coelho, L. D. S. (2023) · Sensors · doi:10.3390/s23063202 · Cited by 90 · cited 1×
“The EWT is a signal processing technique that decomposes a signal into oscillatory modes with different scales and frequencies [63].”
MethodSelf-citationXu, Y., Deng, Y., Zhao, J., Tian, W., & Ma, C. (2019) · IEEE Transactions on Instrumentation and Measurement · doi:10.1109/tim.2019.2928534 · Cited by 85
Zhang, X., Wang, J., Liu, Z., & Wang, J. (2018) · ISA Transactions · doi:10.1016/j.isatra.2018.09.022 · Cited by 83 · cited 1×
“[57] presented a weak feature augmentation method based on EWT and improved adaptive bistable stochastic resonance (IABSR) to extract faulty features in precision machinery.”
MethodWavelet LSTM for Fault Forecasting in Electrical Power Grids
Branco, N. W., Cavalca, M. S. M., Stefenon, S. F., & Leithardt, V. R. Q. (2022) · Sensors · doi:10.3390/s22218323 · Cited by 83 · cited 1×
“Furthermore, using time series forecasting, the increase in the number of failures can be monitored to assess the condition of the system being monitored [23].”
MethodSelf-citationKordestani, H., & Zhang, C. (2020) · Sensors · doi:10.3390/s20071983 · Cited by 78 · cited 1×
“The Savitzky–Golay filter is a polynomial smoothing filter often used to remove noise from time series data while preserving the underlying trends in the data [68].”
Prediction of CO concentration in different conditions based on Gaussian-TCN
Ni, S., Jia, P., Yang, X., et al. (2022) · Sensors and Actuators B Chemical · doi:10.1016/j.snb.2022.133010 · Cited by 73
Stefenon, S. F., Freire, R. Z., Meyer, L. H., Corso, M. P., Sartori, A., Nied, A., Klaar, A. C. R., & Yow, K. (2020) · IET Science Measurement & Technology · doi:10.1049/iet-smt.2020.0083 · Cited by 71 · cited 1×
“Hybrid models which combine noise suppression methods such as seasonal decomposition or wavelet transforms with forecasting models have been increasingly employed [30].”
MethodSelf-citationHybrid deep learning for power generation forecasting in active solar trackers
Stefenon, S. F., Kasburg, C., Nied, A., Klaar, A. C. R., Ferreira, F. C. S., & Branco, N. W. (2020) · IET Generation Transmission & Distribution · doi:10.1049/iet-gtd.2020.0814 · Cited by 64 · cited 1×
“An important observation to consider is that the filters cannot be too coarse to reduce all the variation in the signal, so a proper case evaluation must be performed [32].”
BackgroundSelf-citationAn improved empirical wavelet transform and sensitive components selecting method for bearing fault
Liu, Q., Yang, J., & Zhang, K. (2021) · Measurement · doi:10.1016/j.measurement.2021.110348 · Cited by 62 · cited 1×
“[61] developed an approach to enhance EWT to address the spectrum segmentation flaw and improve the method’s capacity to extract bearing fault data.”
MethodWavelet-based Synchroextracting Transform: An effective TFA tool for machinery fault diagnosis
Shi, Z., Yang, X., Li, Y., & Yu, G. (2021) · Control Engineering Practice · doi:10.1016/j.conengprac.2021.104884 · Cited by 60 · cited 1×
“[58] developed a wavelet-based technique, dubbed wavelet-based synchro extracting transform (WSET), and applied it to fault diagnosis.”
MethodNoise-robust machinery fault diagnosis based on self-attention mechanism in wavelet domain
Tian, A., Zhang, Y., Ma, C., Chen, H., Sheng, W., & Zhou, S. (2022) · Measurement · doi:10.1016/j.measurement.2022.112327 · Cited by 58
Photovoltaic power forecasting using wavelet Neuro-Fuzzy for active solar trackers
Stefenon, S. F., Kasburg, C., Freire, R. Z., Ferreira, F. C. S., Bertol, D. W., & Nied, A. (2020) · Journal of Intelligent & Fuzzy Systems · doi:10.3233/jifs-201279 · Cited by 54 · cited 1×
“[50], who focused on time series forecasting to propose a model and assess solar prediction capability.”
BackgroundSelf-citationVan, M., Hoang, D.-T., & Kang, H. (2020) · Sensors · doi:10.3390/s20123422 · Cited by 53 · cited 1×
“[54] proposed the particle swarm optimization least-squares wavelet support vector machine classifier.”
MethodMedeiros, A., Sartori, A., Stefenon, S. F., Meyer, L. H., & Nied, A. (2021) · Journal of Intelligent & Fuzzy Systems · doi:10.3233/jifs-211126 · Cited by 53 · cited 1×
“However, choosing the appropriate model is a challenging task [26], considering that the characteristics of the data influence model performance and since some methods have specific properties that can be helpful for non-linear forecasting.”
BackgroundSelf-citationDeng, C., Deng, Z., Lu, S. F., He, M., Miao, J., & Peng, Y. (2023) · Sensors · doi:10.3390/s23052542 · Cited by 51 · cited 1×
“For improved signal analysis with non-linearity, techniques such as seasonality decomposition [27], wavelet transform [28], and empirical wavelet transform [29] show promise for denoising.”
MethodFault diagnosis of mine asynchronous motor based on MEEMD energy entropy and ANN
Yang, Z., Kong, C., Wang, Y., Rong, X., & Wei, L. (2021) · Computers & Electrical Engineering · doi:10.1016/j.compeleceng.2021.107070 · Cited by 48 · cited 1×
“[25] proposed an ensemble empirical mode decomposition (EEMD) for the fault diagnosis of asynchronous machines, showing that their approach had a recognition rate of 99%, considering broken rotor bars, air gap eccentricity, and normal state.”
BackgroundLong short-term memory stacking model to predict the number of cases and deaths caused by COVID-19
Fernandes, F., Stefenon, S. F., Seman, L. O., Nied, A., Ferreira, F. C. S., Subtil, M. C. M., Klaar, A. C. R., & Leithardt, V. R. Q. (2021) · Journal of Intelligent & Fuzzy Systems · doi:10.3233/jifs-212788 · Cited by 47 · cited 1×
“Among the time series forecasting models, there are several approaches such as neurofuzzy systems [44], autoregressive integrated moving average (ARIMA) [45], LSTM [46], ensemble learning methods [47], and TFT [48].”
MethodSelf-citationStefenon, S. F., Seman, L. O., Neto, C. S. F., Nied, A., Seganfredo, D. M., da Luz, F. G., Sabino, P. H., González, J. T., & Leithardt, V. R. Q. (2020) · Electronics · doi:10.3390/electronics9111975 · Cited by 47 · cited 1×
“Over the last few years, there has been much attention to applying predictive maintenance methods to predict electrical machine breakdowns by locating anomalies [2].”
BackgroundSelf-citationHasan, F., & Huang, H. (2023) · Sensors · doi:10.3390/s23010530 · Cited by 45
Zhang, X., Kuenzel, S., Colombo, N., & Watkins, C. (2022) · Journal of Modern Power Systems and Clean Energy · doi:10.35833/mpce.2021.000276 · Cited by 43 · cited 1×
“The application of EWT for failure diagnosis extends to other types of machines, such as wind turbines [42], and other forecasting applications [43].”
BackgroundToma, R. N., Gao, Y., Piltan, F., Im, K., Shon, D., Yoon, T. H., Yoo, D.-S., & Kim, J.-M. (2022) · Sensors · doi:10.3390/s22228958 · Cited by 42 · cited 1×
“[53] proposed a system for classifying bearing faults.”
BackgroundDiagnosis of Multiple Faults in Rotating Machinery Using Ensemble Learning
Inyang, U. I., Petrunin, I. V., & Jennions, I. K. (2023) · Sensors · doi:10.3390/s23021005 · Cited by 38
Temperature Prediction of Seasonal Frozen Subgrades Based on CEEMDAN-LSTM Hybrid Model
Chen, L., Liu, X., Zeng, C., He, X., Chen, F., & Zhu, B. (2022) · Sensors · doi:10.3390/s22155742 · Cited by 37 · cited 1×
“Real-world time series data generally exhibit non-linearities [12], making it challenging to apply conventional prediction techniques [13].”
BackgroundKizito, R., Scruggs, P., Li, X., DeVinney, M., Jansen, J., & Kress, R. L. (2021) · IEEE Access · doi:10.1109/access.2021.3077192 · Cited by 33 · cited 1×
“Given the potential for significant reductions in maintenance costs, increased productivity, and reduced downtime, predictive maintenance of electrical machinery has become a top priority for companies [1].”
BackgroundFrequency Phase Space Empirical Wavelet Transform for Rolling Bearings Fault Diagnosis
Huang, X., Wen, G., Liang, L., Zhang, Z., & Tan, Y. (2019) · IEEE Access · doi:10.1109/access.2019.2922248 · Cited by 31 · cited 1×
“[41] applied the EWT for machine bearing fault detection.”
MethodPredictive Maintenance of Norwegian Road Network Using Deep Learning Models
Hassan, M. U., Steinnes, O.-M. H., Gustafsson, E. G., Løken, S., & Hameed, I. A. (2023) · Sensors · doi:10.3390/s23062935 · Cited by 31
Chui, K. T., Gupta, B. B., Liu, R. W., & Vasant, P. (2021) · Sensors · doi:10.3390/s21093133 · Cited by 30 · cited 1×
“[34] showed that the signal-to-noise ratio could be improved using an optimized ensemble empirical model combined with WPT.”
MethodZhao, M., Shi, P., Xu, X., Xu, X., Liu, W., & Yang, H. (2022) · Sensors · doi:10.3390/s22187089 · Cited by 25
Ding, J. (2021) · Measurement · doi:10.1016/j.measurement.2021.109135 · Cited by 24 · cited 1×
“In this sense, Ding [59] proposed a double impulsiveness measurement indices bilaterally driven EWT method to detect and diagnose defects.”
MethodFaysal, A., Ngui, W. K., Lim, M., & Leong, M. (2021) · Italian National Conference on Sensors · doi:10.3390/s21238114 · Cited by 23 · cited 1×
“[33] proposed a noise-eliminated ensemble empirical mode decomposition (NEEEMD) method for fault diagnosis in rotating machinery.”
MethodZheng, J., Huang, S., Pan, H., & Jiang, K. (2019) · IEEE Access · doi:10.1109/access.2019.2940627 · Cited by 23 · cited 1×
“Given an input signal x(t) and a mother wavelet ψ(t), the EWT first generates a set of n non-linear and non-stationary functions called intrinsic mode functions (IMFs) using Algorithm 1 [64,65].”
MethodThe Improved Empirical Wavelet Transform and Applications to Seismic Reflection Data
Liu, N., Li, Z., Sun, F., Wang, Q., & Gao, J. (2019) · IEEE Geoscience and Remote Sensing Letters · doi:10.1109/lgrs.2019.2911092 · Cited by 23 · cited 1×
“Given an input signal x(t) and a mother wavelet ψ(t), the EWT first generates a set of n non-linear and non-stationary functions called intrinsic mode functions (IMFs) using Algorithm 1 [64,65].”
MethodXu, Y., Zhang, K., Ma, C., Sheng, Z., & Shen, H. (2019) · IEEE Access · doi:10.1109/access.2019.2902645 · Cited by 21 · cited 1×
“[37] applied the EWT considering an adaptive spectrum segmentation for the improvement in signal processing and fault diagnosis.”
MethodXin, Y., Li, S., Zhang, Z., An, Z., & Wang, J. (2019) · IEEE Access · doi:10.1109/access.2019.2917042 · Cited by 20 · cited 1×
“[38] for rotating machinery and by Xu et al.”
MethodLee, H., & Lee, J. (2023) · Sensors · doi:10.3390/s23073453 · Cited by 18 · cited 1×
“Therefore, advanced techniques were employed to address this issue, including convolutional neural network (CNN) [14], long short-term memory (LSTM) attention [15], and quantile regression [11], to accurately predict machine failures and manage uncertainties present in the data.”
MethodZhao, B., Li, Q., Lv, Q., & Si, X. (2021) · IEEE Access · doi:10.1109/access.2021.3099500 · Cited by 16
Feng, G., Zhang, L., Ai, F., Zhang, Y., & Hou, Y. (2022) · Entropy · doi:10.3390/e24081111 · Cited by 15 · cited 1×
“[52] used an enhanced TFT prediction model to supply air temperatures in high-speed train carriages.”
MethodPassive Fingerprinting of Same-Model Electrical Devices by Current Consumption
Ronkin, M., & Bykhovsky, D. (2023) · Sensors · doi:10.3390/s23010533 · Cited by 8 · cited 1×
“For improved signal analysis with non-linearity, techniques such as seasonality decomposition [27], wavelet transform [28], and empirical wavelet transform [29] show promise for denoising.”
Method
Metadata
- Authors
- Borré, A., Seman, L. O., Camponogara, E., Stefenon, S. F., Mariani, V. C., & Coelho, L. D. S.
- Venue
- Sensors
- Publisher
- MDPI AG
- Published
- 2023-05-05
- Volume
- 23
- Issue
- 9
- Pages
- 4512
- ISSN
- 1424-8220 (ISSN-L)
- OpenAlex
- W4375866544
- Cited by
- 162
- References
- 61
- FWCI
- 17.61
- Type
- journal article
- Language
- en
- Open access
- gold
About these checks
Registry, retraction, Cites and statistics results are computed, not judged, with one exception: a printed reference that no index lists is read once by a model before it counts as a citation error. Claims are extracted by a model and kept only when their quote appears verbatim in the text; each cited source is then checked, with a model’s help, for whether it says what the claim relies on. Those verdicts can be wrong — read the claim’s quote and the cited source before relying on one.
Statistics and arithmetic are recomputed, not judged.
Confidence bands are shown only for fields calibrated on our annotated test set. None is calibrated yet, so no band is shown.
What counts as a claim ›How we measure confidence ›How we check this ›
Are you an author, or do you see a problem with this page? Request a review
Data sources
Bibliographic data: Semantic Scholar, OpenAlex, Crossref.
Contains information from the Semantic Scholar Academic Graph Datasets, made available under the ODC Attribution License (ODC-BY 1.0).
Semantic Scholar’s data is described in: Kinney, R., et al. (2023). The Semantic Scholar Open Data Platform. arXiv:2301.10140.