An Efficient Method for Monitoring Degradation and Predicting the Remaining Useful Life of Mechanical Rotating Components

Yang, Jingli; Yin, Shuangyan; Chang, Yongqi; Gao, Tianyu; Yang, Cheng

doi:10.1109/tim.2020.3032218

Cited by 17 publications

(11 citation statements)

References 32 publications

Supporting

Mentioning

Contrasting

Order By: Relevance

“…During the training phase, we adopt the Adam optimizer with a learning rate of 0.001 and set 100 epochs. Since the size of the first layer-wide convolution kernel has a great influence on the fault discrimination accuracy of the model, our thesis introduces parameter preference to determine the size of the convolutional kernels [29]. The average value of the accuracy is adopted as the metric.…”

Section: Parameter Settingsmentioning

confidence: 99%

An effective method for fault diagnosis of rotating machinery under noisy environment

Xu,

2023

Meas. Sci. Technol.

View full text Add to dashboard Cite

Rotating machinery is widely utilized as mechanical equipment in the industrial field. However, due to the complex working conditions, the existing fault diagnosis methods have failed to address good results in practical applications. To improve the fault diagnosis performance of rotating machinery in a noisy environment, a new multi-scale convolution neural network (MSCN) based on a self-calibrating attention module is proposed. First, this thesis constructs a multi-scale convolution layer with a wide convolution kernel to form an efficient sampling structure at the filter level, which can filter out incoherent noise from the signals and extract rich features. Second, a multi-scale self-calibrating attention module (MS-SCAM) is implemented with two identical self-calibrating convolutional networks (SCN) to continuously focus on significant embeddings and adaptively combine information from different spatial dimensions. Third, the multi-dimensional characteristics are integrated by the feature cascade layer, and then the fault modes are identified though the classifier layer under noise. Finally, based on the Case Western Reserve University (CWRU) datasets and Paderborn University (PU) bearing datasets, the experimental results show that our proposed MSCN can significantly enhance the fault identification ability to rotate machinery in a noisy environment.

show abstract

Section: Parameter Settingsmentioning

confidence: 99%

An effective method for fault diagnosis of rotating machinery under noisy environment

Xu,

2023

Meas. Sci. Technol.

View full text Add to dashboard Cite

show abstract

“…During the actual feature selection process, the selected feature subset may ideally satisfy the aforementioned characteristics simultaneously [33][34][35]. The weighted calculation is the most widely used feature selection criterion, as it provides a comprehensive performance evaluation of the feature indicators [33,35]. However, the current research has some limitations: the existing selection process uses measured data based on the degradation process, which is difficult to apply to cases without historical data.…”

Section: Introductionmentioning

confidence: 99%

A novel performance degradation assessment method for rotating machinery based on the fault information and the dynamic simulation

Zhang,

Qin

et al. 2024

Meas. Sci. Technol.

View full text Add to dashboard Cite

The performance degradation assessment (PDA) of key components such as gears and rolling bearings is the core technology of prognostics and health management for rotating machinery. Most existing PDA methods may have two deficiencies: (1) the assessment indicator constructed does not consider capturing fault characteristics directly; (2) feature selection is generally based on the measured data of different fault levels, which is difficult to obtain in actual processes; moreover, the selection results lack universality and are difficult to extend to other equipment. To address these issues, this paper proposes a novel PDA method based on fault information and dynamic simulation. First, anomaly detection is performed using four well-known indicators in combination with Mahalanobis distance. Secondly, fault identification is performed using envelope spectrum analysis on anomaly signals to determine the fault type, e.g., gear fault and outer race fault. Thirdly, based on the fault type information, the candidate feature set including fault-domain indicators is selected based on the established dynamic simulation signals to obtain a preliminary assessment vector for the first stage. The stability of the fault domain indicators which capture fault characteristics directly is tested through actual measured normal data. It is used as the second stage of selecting to obtain the assessment vector. Finally, the PDA indicator is calculated based on the assessment vector and Mahalanobis distance. Four experiment case studies demonstrate the proposed PDA method can effectively isolate faults with different defect sizes as well as track the whole performance degradation. The above analysis indicates that the proposed PDA method is expected to be used for the actual rotating machinery.

show abstract

“…Model‐based approaches include an exponential model, 10–11 an exponential with extended Kalman filter (EKF) 12 or particle filtering (PF), 13–16 Gaussian process regression (GPR), 17 and log‐linear recursive least‐squares (LL‐RLS) 18 . Data‐driven approaches include convolutional neural network (CNN) 19 and machine learning algorithm, 20 2D‐CNN, 21 deep belief network (DBF), 22 long short‐term model (LSTM), 23 LSTM with encoder‐decoder, 24 recurrent neural network (RNN), 25 restricted Boltzmann machine (RBM), 26 support vector regression (SVR), 27–28 convolutional neural network‐gated recurrent unit (CNN‐GRU), 29 and sparse representation model 30–31 . Hybrid approaches can be found in literature 16,32–36 …”

Section: Introductionmentioning

confidence: 99%

Transfer learning based deep learning model and control chart for bearing useful life prediction

Wang

Gomez

Amogne

et al. 2023

Quality & Reliability Eng

View full text Add to dashboard Cite

The remaining useful life (RUL) of the machine is one of the key information for predictive maintenance. If there is a lack of predictive maintenance strategy, it will increase the maintenance and breakdown costs of the machine. We apply transfer learning techniques to develop a new method that predicts the RUL of target data using degradation trends learned from complete bearing test data called source data. The training length of the model plays a crucial role in RUL prediction. First, the exponentially weighted moving average (EWMA) chart is used to identify the abnormal points of the bearing to determine the starting point of the model's training. Secondly, we propose transfer learning based on a bidirectional long and short-term memory with attention mechanism (BiLSTMAM) model to estimate the RUL of the ball bearing. At the same time, the public data set is used to compare the estimation effect of the BiLSTMAM model with some published models. The BiLSTMAM model with the EWMA chart can achieve a score of 0.6702 for 11 target bearings. The accuracy of the RUL estimation ensures a reliable maintenance strategy to reduce unpredictable failures.

show abstract

An Efficient Method for Monitoring Degradation and Predicting the Remaining Useful Life of Mechanical Rotating Components

Cited by 17 publications

References 32 publications

An effective method for fault diagnosis of rotating machinery under noisy environment

An effective method for fault diagnosis of rotating machinery under noisy environment

A novel performance degradation assessment method for rotating machinery based on the fault information and the dynamic simulation

Transfer learning based deep learning model and control chart for bearing useful life prediction

Contact Info

Product

Resources

About