Figure - available from: Advances in Mechanical Engineering
This content is subject to copyright.
Source publication
The fault diagnosis of rolling element bearings is very important for ensuring the safe operation of rotary machineries. Targeting the nonstationary characteristics of the vibration signals of rolling element bearings, a novel approach based on dual-tree complex wavelet packet transform, improved intrinsic time-scale decomposition, and the online s...
Similar publications
An SVM-based recognition method for the safety of oil and gas pipeline was proposed due to limitation of the traditional learning methods based on empirical risk minimization. The vibration signals along the pipelines are obtained with the distributed optical fiber vibration sensor on the basis of Mach-Zehnder optical fiber interferometer theory. T...
Citations
... In their study, Gao et al. [74] applied an integrated extreme learning machine (IELM) to detect mechanical issues specifically in high-voltage circuit breakers. Tang et al. [75] successfully diagnosed bearing faults by utilizing a sequential extreme learning machine. Chen et al. [76] employed a summation wavelet extreme value machine (SW-ELM) to effectively assess different types of faults. ...
In the field of in-process monitoring for the production of safety-critical parts, the utilization of Machine Learning (ML) models has demonstrated promising potential for enhancing the manufacturing process. Specifically, Supervised Classification Algorithms, selected based on the complexity of the problem, can significantly improve the performance of ML models. While ML primarily relies on training data to comprehend input-output relationships, its ability to draw rapid conclusions is noteworthy. Deep neural networks have emerged as valuable tools, offering more accurate analytical data compared to Finite Element Method (FEM) models. They present a cost-effective alternative to empirical data by incorporating error compensation in a closed loop. Nevertheless, vision-based techniques, such as those used in this study, necessitate high-quality input images and demand substantial computational power for image classification. To address these challenges, this study proposes the use of a shallow Convolutional Neural Network (CNN)-based architecture for detecting in-process defects in parts manufactured using Additive Manufacturing (AM) technology. Complementing this, Support Vector Machine (SVM) and Extreme Learning Machine (ELM) algorithms are integrated with the CNN architecture. Additionally, the research takes into account the high costs associated with data collection and installation, data unavailability, challenges in data labeling, as well as common difficulties like overfitting and underfitting of ML models. These factors often pose constraints on the application of ML solutions within the context of AM. The study aims to address these issues and shed light on the potential of ML for AM applications. By employing a combination of CNN, SVM, and ELM algorithms, this research delves into the effectiveness of ML models in defect detection during the AM process. The insights derived from this study contribute to mitigating the aforementioned limitations and pave the way for a broader adoption of ML solutions in AM. This optimization aims to enhance performance, reduce costs, and improve the overall quality of manufactured parts.
... Time-frequency analysis based on the intrinsic time-scale decomposition can quantitatively describe the relationship between frequency and time, accurately analyzing time-varying signals [10]. On the basis of these advantages, scholars introduced this method from the medical field to the fault diagnosis of mechanical signals [11][12][13][14][15][16][17][18][19][20][21][22]. For example, Lin and Chang published a rolling-bearing fault diagnosis method based on an enhanced kurtosis spectrum and intrinsic time-scale decomposition [11]. ...
... Hu and Xiang et al. proposed ensemble intrinsic time-scale decomposition to the fault diagnosis of fan gear [15]. Tong, Cao, et al. proposed improved intrinsic time-scale decomposition combined with complex tree wavelet packet transform and singular-value decomposition and used it to diagnose rolling-bearing faults [16]. Liu and Zhang et al. proposed the use of intrinsic time-scale decomposition for diesel-engine fault diagnosis [17]. ...
The early fault diagnosis of rolling bearings has always been a difficult problem due to the interference of strong noise. This paper proposes a new method of early fault diagnosis for rolling bearings with entropy participation. First, a new signal decomposition method is proposed in this paper: intrinsic time-scale decomposition based on time-varying filtering. It is introduced into the framework of complete ensemble intrinsic time-scale decomposition with adaptive noise (CEITDAN). Compared with traditional intrinsic time-scale decomposition, intrinsic time-scale decomposition based on time-varying filtering can improve frequency-separation performance. It has strong robustness in the presence of noise interference. However, decomposition parameters (the bandwidth threshold and B-spline order) have significant impacts on the decomposition results of this method, and they need to be artificially set. Aiming to address this problem, this paper proposes rolling-bearing fault diagnosis optimization based on an improved coyote optimization algorithm (COA). First, the minimal generalized refined composite multiscale sample entropy parameter was used as the objective function. Through the improved COA algorithm, optimal intrinsic time-scale decomposition parameters based on time-varying filtering that match the input signal are obtained. By analyzing generalized refined composite multiscale sample entropy (GRCMSE), whether the mode component is dominated by the fault signal is determined. The signal is reconstructed and decomposed again. Finally, the mode component with the highest energy in the central frequency band is selected for envelope spectrum variation for fault diagnosis. Lastly, simulated and experimental signals were used to verify the effectiveness of the proposed method.
... Hu and Xiang proposed the application of the ensemble intrinsic time-scale decomposition of fan gear [17]. Tong and Cao et al. proposed an improved intrinsic time-scale decomposition with wavelet packet transform and singular-value decomposition to perform fault diagnosis on rolling bearings [18]. Liu and Zhang et al. proposed using intrinsic time-scale decomposition to diagnose diesel-engine faults [19]. ...
When early failures in rolling bearings occur, we need to be able to extract weak fault characteristic frequencies under the influence of strong noise and then perform fault diagnosis. Therefore, a new method is proposed: complete ensemble intrinsic time-scale decomposition with adaptive Lévy noise (CEITDALN). This method solves the problem of the traditional complete ensemble intrinsic time-scale decomposition with adaptive noise (CEITDAN) method not being able to filter nonwhite noise in measured vibration signal noise. Therefore, in the method proposed in this paper, a noise model in the form of parameter-adjusted noise is used to replace traditional white noise. We used an optimization algorithm to adaptively adjust the model parameters, reducing the impact of nonwhite noise on the feature frequency extraction. The experimental results for the simulation and vibration signals of rolling bearings showed that the CEITDALN method could extract weak fault features more effectively than traditional methods.
... Besides, DTCWPT utilizes two parallel and independent discrete wavelet packets to decompose the low-frequency and high-frequency parts, exhibiting extremely high resolution while also effectively suppressing the frequency aliasing phenomenon. e decomposition and reconstruction of DTCWPT [32] are presented in Figure 1. ...
The health condition of rolling bearings, as a widely used part in rotating machineries, directly influences the working efficiency of the equipment. Consequently, timely detection and judgment of the current working status of the bearing is the key to improving productivity. This paper proposes an integrated fault identification technology for rolling bearings, which contains two parts: the fault predetection and the fault recognition. In the part of fault predetection, the threshold based on amplitude-aware permutation entropy (AAPE) is defined to judge whether the bearing currently has a fault. If there is a fault in the bearing, the fault feature is adequately extracted using the feature extraction method combined with dual-tree complex wavelet packet transform (DTCWPT) and generalized composite multiscale amplitude-aware permutation entropy (GCMAAPE). Firstly, the method decomposes the fault vibration signal into a set of subband components through the DTCWPT with good time-frequency decomposing capability. Secondly, the GCMAAPE values of each subband component are computed to generate the initial candidate feature. Next, a low-dimensional feature sample is established using the t-distributed stochastic neighbor embedding (t-SNE) with good nonlinear dimensionality reduction performance to choose sensitive features from the initial high-dimensional features. Afterwards, the featured specimen representing fault information is fed into the deep belief network (DBN) model to judge the fault type. In the end, the superiority of the proposed solution is verified by analyzing the collected experimental data. Detection and classification experiments indicate that the proposed solution can not only accurately detect whether there is a fault but also effectively determine the fault type of the bearing. Besides, this solution can judge the different faults more accurately compared with other ordinary methods.
... This has the advantage of being approximately shift invariant, as opposed to the shift variant WPT. Among several other applications, DTCWPT has been used for diagnosis of mechanical faults in gearbox and roller bearings in [16] and [17]. In this paper, we employ DTCWPT to extract features and use an SVM-based classifier to diagnose the condition of the motor. ...
Fault detection and identification (FDI) of electrical motors is crucial to ensuring smooth operation of several industrial systems. Detection and diagnosis of faults at the incipient stage allows corrective actions to be adopted so as to curb the severity of faults. However, FDI of incipient faults has proved to be elusive to traditional methods of fault diagnosis. With recent developments in statistical machine learning, new methods are proposed that can be used for FDI. In this article we adopt three tools (support vector machine, convolutional and recurrent networks) from machine learning to address the challenge of FDI of incipient faults. We perform FDI of a DC motor with the most commonly and readily measured current data. Results from experimental data reveal that the convolutional network performs the best of the three methods. A comparative study of the performance of the three methods under different types of operating conditions is provided. Sensitivity of the techniques to noise in measurements is also studied. The proposed approach serves as a reliable tool for FDI of DC motor under different types of loading conditions.
... Available publications often deal with the diagnosing the condition of rolling bearings based on the measurement of generated vibrations [1], [2], [3], [4]. Equally often, we can find articles focusing on tests of vibration sensors, used to measure bearing vibrations [5], [6], [7]. ...
In this paper, a novel bearing fault diagnosis approach based on pseudo fourth-order moment is proposed and verified. Based on the wavelet decomposition of different vibration signals, the pseudo fourth-order moment is calculated, and a new fault feature–the feature angle is proposed. Feature angle between these pseudo fourth-order moments is obtained, and corresponding working condition maybe characterized by this feature angle. The ranges of these feature angles are determined by simulation experiments with a large amount of data. An assessment index (namely selection index of optimal data size) is constructed to select the optimal amount of computational data. Meanwhile, extreme learning machine (ELM) model is established to classify different working conditions. According to the ELM model obtained from training data, the accuracy of test data classification reaches 95.42%, which proves the effectiveness of present bearing fault diagnosis method.
Extreme learning machine (ELM) has better operation efficiency in fault diagnosis. However, the recognition accuracy of ELM algorithm is actually affected by the activation function. Moreover, most of the testing dataset are coming from high precision and expensive sensors. In this paper, raw data are collected by a low-cost attitude sensor, which is installed on the mobile platform of a delta 3D printer. A doublet activation function is proposed to improve the performance of ELM, named doublet ELM (DELM). The proposed method is evaluated using experimental data collected from the 3D printer, and its advantages are demonstrated by comparing with other activation functions. The experimental results indicate that the proposed method leads to the highest accuracy in different hidden nodes and the testing classification rate achieves 93% and 96% using only 8.33% of the dataset for model training, for R75 and R90 sub-datasets, respectively. Moreover, compared with peer methods, such as random forest, echo state network, and so on, the results show that the present DELM exhibits the best performance in small-sample and improves the accuracy of the 3D printer fault diagnosis.
Within the framework of ensuring availability at oil and gas processing facilities, an analysis of methods for assessing the reliability of a technical object based on the reliability, availability and maintainability of individual elements was made. The difficulties arising from the complex assessment of the reliability of complex technical objects using different methods are shown: A number of methods did not allow to assess the entire complexity of the object, and other methods led to an increase in the complexity of calculations with an increase in the number of individual elements. The authors propose to use combinations of previously known methods at different hierarchical levels for system analysis. An algorithm for assessing reliability based on dividing a complex object into elements, the evaluation of the reliability of which is determined by one of the most suitable methods, such as the Markov models of states and transitions or statistical models, has been developed. Additional designations are proposed for the unambiguous interpretation and structuring of the reliability assessment system. As an example, the calculation of the failure-free operation of the gas treatment unit of the tar visbreaker was made. The possibility of calculating complex interdependent systems, where linear statistical calculation methods are not applicable, and the labor intensity for the Markov method has power-law dependence, is shown.
Background
Since strong background noise is inevitably embedded in non-stationary signals collected under faulty condition, and the difficulty to obtain fault characteristic frequencies in practice, noise reduction should be paid more attention to extract fault features.
Purpose
A novel fault feature- extracted method is put forward by means of dual-tree complex wavelet transform (DTCWT) and singular value decomposition (SVD).
Methods
Dual-tree complex wavelet transform is considered as the key technique to perform multilevel decomposition and several different frequency band components are obtained. A Hankel matrix is constructed by employing the components which contain the fault information, and the singular values are obtained after the singular value decomposition. According to principal component analysis, the number of singular values is determined to realize noise reduction using SVD reconstruction. Finally, the fault frequency can be identified accurately by Hilbert envelop spectrum.
Results
The results of the experiments and practical engineering applications demonstrate that the fault feature of wind turbine can be extracted effectively.
Conclusions
The integrated method of DTCWT and SVD is adopted to reduce noise, which can provide some guidance for theoretical research and engineering applications in fault diagnosis.
