Search

This paper presents a fault diagnosis method for a vacuum contactor using the generalized Stockwell transform (GST) of vibration signals. The objective is to solve the problem of low diagnostic performance efficiency caused by the inadequate feature extraction capability and the redundant pixels in the graph background. The proposed method is based on the time-frequency graph optimization technique and ShuffleNetV2 network. Firstly, vibration signals in different states are collected and converted into GST time-frequency graphs. Secondly, multi-resolution GST time-frequency graphs are generated to cover signal characteristics in all frequency bands by adjusting the GST Gaussian window width factor λ. The OTSU algorithm is then combined to crop the energy concentration area, and the size of these time-frequency graphs is optimized by 68.86%. Finally, considering the advantages of the channel split and channel shuffle methods, the ShuffleNetV2 network is adopted to improve the feature learning ability and identify fault categories. In this paper, the CKJ5-400/1140 vacuum contactor is taken as the test object. The fault recognition accuracy reaches 99.74%, and the single iteration time of model training is reduced by 19.42%. [ABSTRACT FROM AUTHOR]

The health of mechanical components can be assessed by analyzing the vibration and acoustic signals they produce. These signals contain valuable information about the component's condition, often encoded within specific frequency bands. However, extracting this information is challenging due to noise contamination from various sources. Narrow-band amplitude demodulation presents a robust technique for isolating fault-related information within the signal. This work proposes a novel approach based on cluster-based segmentation for demodulating the signal and extracting the frequency band of interest. The segmentation process leverages the criteria of maximum L-kurtosis and minimum entropy. L-kurtosis maximizes impulsiveness in the signal, while minimum entropy signifies a low degree of randomness and high cyclo-stationarity, and both characteristics are crucial for identifying the desired frequency band. Simulations and experimental tests using vibration signals from different gears demonstrate the effectiveness of this technique. The processed envelope of the signal exhibits distinct improvements, highlighting the ability to accurately extract the fault-related information embedded within the complex noise-ridden signals. This approach offers a promising solution for accurate and efficient fault diagnosis in mechanical systems, contributing to enhanced reliability and reduced downtime. [ABSTRACT FROM AUTHOR]

In single crystal diamond tool grinding, cutting-edge quality is affected by various parameters, with tool vibration playing a crucial role. Due to environmental factors and the complexity of the process, vibration signals are often noisy and non-stationary. This study proposes a denoising method that optimizes the wavelet threshold function using an Improved Genetic Algorithm (IGA). The method introduces a configurable value α in the wavelet threshold function, which is optimized with IGA to improve denoising. MATLAB R2018b simulations show that this approach achieves better denoising, a higher signal-to-noise ratio, and lower mean square error. [ABSTRACT FROM AUTHOR]

The present study emphasizes an optimized deep learning algorithm for gearbox fault detection using vibration, sound, and acoustic emission signals. Statistical and acoustic features are extracted from these signals, and various neural network algorithms are explored. The supervised deep feed forward neural network (DFFNN) demonstrates excellent performance with vibration signals but limited accuracy with sound and acoustic emission signals. To address this, unsupervised algorithms are optimized and compared with vibration-based classification. The findings show that unsupervised neural networks, particularly the auto-encoder and stacked auto-encoder architectures, achieve improved classification accuracy by leveraging the unique characteristics of acoustic emission signals. The unsupervised models also effectively overcome the vanishing gradient problem via regularization, enhancing their training efficiency. The stacked auto-encoder, with multiple layers of encoders and decoders, reduces computation time by 40% and memory consumption. These optimized algorithms hold promise for automated fault detection systems. The auto-encoder and stacked auto-encoder, utilizing vibration, sound, and acoustic emission signals, offer enhanced classification accuracy and can facilitate real-time monitoring of rotating mechanical systems. However, further optimization is needed to maximize their performance. In a nutshell, the supervised DFFNN excels in utilizing vibration signals for fault detection, while the unsupervised models exploit the distinctive characteristics of acoustic emission signals. Future research will focus on refining these algorithms to enhance their effectiveness. Implementing these optimized deep learning approaches can lead to autonomous fault detection systems, eliminating the need for continuous human supervision. [ABSTRACT FROM AUTHOR]

A mechanical vibration fault diagnosis is a key means of ensuring the safe and stable operation of transformers. To achieve an accurate diagnosis of transformer vibration faults, this paper proposes a novel fault diagnosis method based on time-shift multiscale increment entropy (TSMIE) combined with CatBoost. Firstly, inspired by the concept of a time shift, TSMIE was proposed. TSMIE effectively solves the problem of the information loss caused by the coarse-graining process of traditional multiscale entropy. Secondly, the TSMIE of transformer vibration signals under different operating conditions was extracted as fault features. Finally, the features were sent into the CatBoost model for pattern recognition. Compared with different models, the simulation and experimental results showed that the proposed model had a higher diagnostic accuracy and stability, and this provides a new tool for transformer vibration fault diagnoses. [ABSTRACT FROM AUTHOR]

This work examines swipe-based interactions on smart devices, like smartphones and smartwatches, that detect vibration signals through defined swipe surfaces. We investigate how these devices, held in users' hands or worn on their wrists, process vibration signals from swipe interactions and ambient noise using a support vector machine (SVM). The work details the signal processing workflow involving filters, sliding windows, feature vectors, SVM kernels, and ambient noise management. It includes how we separate the vibration signal from a potential swipe surface and ambient noise. We explore both software and human factors influencing the signals: the former includes the computational techniques mentioned, while the latter encompasses swipe orientation, contact, and movement. Our findings show that the SVM classifies swipe surface signals with an accuracy of 69.61% when both devices are used, 97.59% with only the smartphone, and 99.79% with only the smartwatch. However, the classification accuracy drops to about 50% in field user studies simulating real-world conditions such as phone calls, typing, walking, and other undirected movements throughout the day. The decline in performance under these conditions suggests challenges in ambient noise discrimination, which this work discusses, along with potential strategies for improvement in future research. [ABSTRACT FROM AUTHOR]

Changes in operating conditions often cause the distribution of signal features to shift during the bearing fault diagnosis process, which will result in reduced diagnostic accuracy of the model. Therefore, this paper proposes a dual-channel parallel adversarial network (DPAN) based on vision transformer, which extracts features from acoustic and vibration signals through parallel networks and enhances feature robustness through adversarial training during the feature fusion process. In addition, the Wasserstein distance is used to reduce domain differences in the fused features, thereby enhancing the network's generalization ability. Two sets of bearing fault diagnosis experiments were conducted to validate the effectiveness of the proposed method. The experimental results show that the proposed method achieves higher diagnostic accuracy compared to other methods. The diagnostic accuracy of the proposed method can exceed 98%. [ABSTRACT FROM AUTHOR]

The fractal dimension (FD) is an effective indicator to characterize various signals in engineering. However, the FD is nearly twice that of its maximum value when examining high-frequency-dominant signals, such as those in milling chatter. Previous studies in the literature have generally employed signal-pre-processing methods that require a significant amount of time to lower the FD range, thus enabling the distinguishment of different states while disabling online monitoring. A new quantitative method based on the FD within a fixed interval was constructed in this study to address this issue. First, the relationship between the fixed-interval fractal dimension (FFD) and the energy ratio (ER), named the fractal complexity curve (FC-Curve), was established, and the sensitivity region of the FFD was determined. Second, a high-frequency suppression filter (HSF) with a high calculation speed was proposed to suppress the signal's ER so the FFD could be adjusted within its sensitivity region. Moreover, a fast energy ratio (FER) correlated with the FFD was proposed using the FC-Curve and HSF to quantitatively analyze dominant high-frequency signals. Finally, the proposed method was verified via its application in milling chatter identification. The FER method accomplished signal analysis more quickly than the traditional energy ratio difference and entropy methods, demonstrating its feasibility for online monitoring and chatter suppression in practical engineering applications. [ABSTRACT FROM AUTHOR]

Non-stationary fault detection under bearing fault operation of induction motor is investigated in this paper. For this aim, the vibration signal is analyzed by wavelet method and pencil matrix method. The pencil matrix (PM) or (MP) method has been combined with wavelet transform (WT), in order to reconstruct the non-stationary signal and detect the bearing fault frequency. For validation of results, an experimental setup is used for an induction motor under different load operation and with failure on its inner race. The application of the proposed technique on vibration signal under non-stationary state show that fault can be characterized by a particular signature that it is not possible with fast Fourier transform (FFT). [ABSTRACT FROM AUTHOR]

Copyright of Diamond & Abrasives Engineering is the property of Zhengzhou Research Institute for Abrasives & Grinding and its content may not be copied or emailed to multiple sites or posted to a listserv without the copyright holder's express written permission. However, users may print, download, or email articles for individual use. This abstract may be abridged. No warranty is given about the accuracy of the copy. Users should refer to the original published version of the material for the full abstract. (Copyright applies to all Abstracts.)

Objectives: With the development of the manufacturing industry, precision grinding has become an indispensable part of the high-end manufacturing field. As a key tool for precision grinding, the surface reshaping technology of diamond rollers is one of the critical technologies for making diamond rollers. Currently, the diamond grinding wheel grinding method is commonly used for diamond roller precision reshaping. In the process of precision reshaping, the surface profile of the diamond roller is an important indicator for assessing its reshaping quality. However, in the current reshaping detection, the workers often stop the machine to remove the rollers and place them on the profiler, which greatly increases the time and the cost of roller reshaping. Therefore, to improve the efficiency of diamond roller reshaping, a method based on sensors to monitor the state of contour reshaping during the process of diamond roller reshaping is explored. Methods: To realize online monitoring and discrimination of the diamond roller contour reshaping state, the diamond roller grinding monitoring platform was first built according to the structure of the GC-X3 machine tool, and LabVIEW was used to read and save the signals collected by the acquisition card. Secondly, the diamond rollers with different contour reshaping states were processed by the GC-X3 machine tool, and the signals with different reshaping states collected by the computer were analyzed in time and frequency domains. The wavelet packet decomposition method was used to decompose the signals according to the characteristics of time and frequency domains, and the signals of different nodes obtained after decomposition were calculated and compared to identify and determine the characteristic values of the signals. Finally, the machine learning method was used to establish the recognition model of the diamond roller contour reshaping state. The model is based on the mapping relationship between the characteristic values of the vibration signals and the contour reshaping state to realize the intelligent recognition of the contour reshaping state of the diamond roller. Results: After preprocessing the signal, it can be observed through the time and the frequency domains that the maximum amplitudes of the X and Y axis vibration signals decrease with the improvement of the roller profile accuracy. The amplitudes of the X and Y axis vibration signals during rough trimming are 1.00 and 0.50 V, respectively. The amplitudes of the X and Y axis vibration signals during fine trimming are 0.50 and 0.40 V, respectively. The amplitudes of X and Y axis vibration signals at the completion of trimming are 0.29 and 0.27 V respectively, and the main frequencies of signals at the completion of rough, fine, and finishing are all distributed within 1 000 Hz. The root mean square and the variance of wavelet packet coefficients mainly occur at nodes (3,0) to (3,3), and as the grinding accuracy increases, the root mean square and the variance of wavelet packet coefficients decrease gradually. When rough trimming diamond rollers with diamond grinding wheels, the roundness error value of the roller profile is larger than that after fine trimming and the completion of trimming, resulting in a larger local grinding allowance. A large number of abrasive particles from the diamond grinding wheel come into contact with the roller, which increases the material removal amount of the diamond roller, the grinding force also increases, and the resulting vibration also increases accordingly. During the fine trimming process, the roundness error of the roller contour is reduced, so that the diamond wheel cutting the diamond roller allowance is reduced, the removal amount is also reduced, the grinding force is reduced, and the vibration generated is also reduced. At the same time, the root mean square and the variance of the wavelet packet coefficients in the completion of trimming decrease correspondingly as the roundness error value of the roller profile decreases. As the trend and fluctuation of the root mean square and the variance of the wavelet packet coefficients in the X and Y axes are generally consistent, the root mean square and the variance of the wavelet packet coefficients in the direction of the X-axis are selected as the feature vectors for the classification of experimental data. The wavelet packet is used to decompose the X-axis signal, and the variance and the root mean square of each node are calculated to obtain 16 characteristic values of the vibration signal. There is a mapping relationship between these characteristic values and the reshaping state of the roller, and the random forest algorithm is used to classify the vibration signals based on these 16 characteristic values, and the recognition accuracy reaches 93.3%. Conclusions: The contour reshaping of the diamond roller is the key to the production of diamond rollers. In the process of roller machining, the method combining sensor monitoring and machine learning models is used to judge the reshaping state of the diamond roller. The method can be used to identify the modified state of the diamond roller profile, reduce the number of roller disassemblies and roller profile detections, and improve the processing efficiency.

The health status of bearings seriously affects the operational efficiency of equipment, and it is important to carry out bearing health status detection. A bearing fault diagnosis method based on stepwise variational modal decomposition (SVMD) with adaptive initialization center frequency and Gramian angular difference field is proposed. Firstly, a method of center frequency initialization base on frequency energy distribution characteristics is proposed to improve the decomposition speed and stability. Secondly, SVMD with single component decomposition and local decomposition is proposed to improve decomposition efficiency. It can effectively avoid inconsistency in different signal parameter settings and ensures consistency in the number of signal components, which is very suitable for batch processing of signals. Finally, Gramian angular field (GAF) and convolutional neural networks (CNNs) are combined to extract features of the reconstructed signal spectrum and enhance the differential characteristics between different signal spectrum. The experiment shows that the center frequency initialization method can shorten the single decomposition time from 11.13 to 6.71 s. The overall recognition rate can reach 95.2%, which is at least 1.9% higher than other decomposition methods. [ABSTRACT FROM AUTHOR]

The aim of this study is to energy harvesting and modeling from a soft and flexible nanogenerator based on nanofibers (PVDF/ZnO-RGO) on the arm of a multi-rotor UAV. The nanogenerator modeling was done to improve the usability of piezoelectric nanogenerators to provide part of the electric energy required by drones. Tests were conducted on the experimental setup and practical flight conditions. In each test and a period of 30 s, the vibration signals of the UAV arm were stored, and the vibration data extracted in the time domain and the electrical voltage of the nanogenerator were obtained through the vibrations of the drone, and its values were recorded in each experiment. After signal processing, the best vibration features were used for modeling. In the modeling process, some data were considered for training and testing in ANFIS network. The ANFIS inputs included the selected vibration features and the outputs included the stored voltage. The modeling results were analyzed based on increasing the amplitude of vibrations and the duration of exposure of the nanogenerator to vibration led to higher voltage generation. The maximum voltage recorded during the 30-s flight time of the tested UAV was 64 mV at 3.54, 49.8, and 0.48 Hz along the x, y, and z axes, respectively. The results showed that the application of energy harvesting on different vibration systems could be improved through modeling. It can be used to predict the amount of electric energy produced according to the vibrations of the drone in the future.

This paper presents a comprehensive methodology for developing a vibration measurement system, with a primary focus on monitoring rolling element bearings, widely recognised as pivotal components in generic machine vibration monitoring. Through the utilisation of sensors to capture vibration data signals, the system aims to meticulously analyse these signals to discern any indications of bearing health deterioration, thus facilitating the early identification of potential pre-breakdown conditions. The proposed system initiates the process by internally decomposing the three-axis vibration signal, enabling subsequent transmission of processed data to a remote location for the identification of specific faults in the roller element bearings. Installation of a three-axis accelerometer on the bearing housing, positioned radially, facilitates the capture of raw vibration data along one axial and two radial directions. Moreover, efforts to enhance computational efficiency involve cascading multiple identical modules in a serial pipeline structure, allowing for iterative loop computation of intrinsic mode functions within Empirical Mode Decomposition (EMD). Real-time computation is achieved through this serial pipeline structure, necessitating two distinct clock frequencies: one at 12.5 MHz for sampling and storing 3-axis vibration data in ping pong buffers, and another at 40 MHz for reading and processing the data signal retrieved from these buffers. Validation of the proposed method is demonstrated through simulation results obtained via synthesised Verilog code on Xilinx Artix-7 FPGA (XC7A35T-1CPG236C), affirming its efficacy in real-time applications. [ABSTRACT FROM AUTHOR]

The insulated bearing may be damaged when operating under the action of current. Firstly, a novel dynamic model of insulated bearing-steel pedestal system with shaft current damage is proposed, and the model is solved based on the Runge-Kutta algorithm. Then, the model is verified by the experimental results. Based on the solution results of the dynamic model, the influence of mass and damage on the vibration signal is analyzed, the effect of damage on the stability of the insulated bearing-steel pedestal system is studied, and the influence of radial clearance on the vibration characteristics of the insulated bearing-steel pedestal system is discussed. The results indicate that, the vibration acceleration signal of the pedestal shows a positive correlation with the mass mi and the mass mo. As the length of damage increases, a few points begin to leave the quasi-periodic ring, and the quasi-periodic ring in Poincaré map gradually breaks. The coupling effect of shaft current damage and radial clearance makes the insulated bearing-steel pedestal system easier to enter an unstable motion state. [ABSTRACT FROM AUTHOR]

Stirred reactors are key equipment in production, and unpredictable failures will result in significant economic losses and safety issues. Therefore, it is necessary to monitor its health state. To achieve this goal, in this study, five states of the stirred reactor were firstly preset: normal, shaft bending, blade eccentricity, bearing wear, and bolt looseness. Vibration signals along x, y and z axes were collected and analyzed in both the time domain and frequency domain. Secondly, 93 statistical features were extracted and evaluated by ReliefF, Maximal Information Coefficient (MIC) and XGBoost. The above evaluation results were then fused by D-S evidence theory to extract the final 16 features that are most relevant to the state of the stirred reactor. Finally, the CatBoost algorithm was introduced to establish the stirred reactor health monitoring model. The validation results showed that the model achieves 100% accuracy in detecting the fault/normal state of the stirred reactor and 98% accuracy in diagnosing the type of fault. [ABSTRACT FROM AUTHOR]

Monitoring the condition of rotating machinery is critical in aerospace applications like aircraft engines and helicopter rotors. Faults in these components can lead to catastrophic outcomes, making early detection essential. This paper proposes a novel approach using vibration signals and time series prediction methods for fault detection in rotating aerospace machinery. By extracting relevant features from vibration signals and using prediction models, fault severity can be effectively quantified. Our experimental results show that the proposed method has potential in early fault detection and is applicable to various types of bearing faults and the different statuses of these faults under complex running conditions, achieving very good generalization ability. [ABSTRACT FROM AUTHOR]

Copyright of FME Transactions is the property of University of Belgrade, Faculty of Mechanical Engineering and its content may not be copied or emailed to multiple sites or posted to a listserv without the copyright holder's express written permission. However, users may print, download, or email articles for individual use. This abstract may be abridged. No warranty is given about the accuracy of the copy. Users should refer to the original published version of the material for the full abstract. (Copyright applies to all Abstracts.)

To enhance fault detection in slewing bearing vibration signals, an advanced noise-reduction model, HRCSA-VMD-WT, is designed for effective signal noise elimination. This model innovates by refining the Chameleon Swarm Algorithm (CSA) into a more potent Hybrid Reinforcement CSA (HRCSA), incorporating strategies from Chaotic Reverse Learning (CRL), the Whale Optimization Algorithm's (WOA) bubble-net hunting, and the greedy strategy with the Cauchy mutation to diversify the initial population, accelerate convergence, and prevent local optimum entrapment. Furthermore, by optimizing Variate Mode Decomposition (VMD) input parameters with HRCSA, Intrinsic Mode Function (IMF) components are extracted and categorized into noisy and pure signals using cosine similarity. Subsequently, the Wavelet Threshold (WT) denoising targets the noisy IMFs before reconstructing the vibration signal from purified IMFs, achieving significant noise reduction. Comparative experiments demonstrate HRCSA's superiority over Particle Swarm Optimization (PSO), WOA, and Gray Wolf Optimization (GWO) regarding convergence speed and precision. Notably, HRCSA-VMD-WT increases the Signal-to-Noise Ratio (SNR) by a minimum of 74.9% and reduces the Root Mean Square Error (RMSE) by at least 41.2% when compared to both CSA-VMD-WT and Empirical Mode Decomposition with Wavelet Transform (EMD-WT). This study improves fault detection accuracy and efficiency in vibration signals and offers a dependable and effective diagnostic solution for slewing bearing maintenance. [ABSTRACT FROM AUTHOR]

The current paper presents helical gearbox defect detection models built from raw vibration signals measured using a triaxial accelerometer. Gear faults, such as localized pitting, localized wear on helical pinion tooth flanks, and low lubricant level, are under observation for three rotating velocities of the actuator and three load levels at the speed reducer output. The emphasis is on the strong connection between the gear faults and the fundamental meshing frequency GMF, its harmonics, and the sidebands found in the vibration spectrum as an effect of the amplitude modulation (AM) and phase modulation (PM). Several sets of features representing powers on selected frequency bands or/and associated peak amplitudes from the vibration spectrum, and also, for comparison, time-domain and frequency-domain statistical feature sets, are proposed as predictors in the defect detection task. The best performing detection model, with a testing accuracy of 99.73%, is based on SVM (Support Vector Machine) with a cubic kernel, and the features used are the band powers associated with six GMF harmonics and two sideband pairs for all three accelerometer axes, regardless of the rotation velocities and the load levels. [ABSTRACT FROM AUTHOR]

Copyright of SDU Journal of Engineering Sciences & Design / Mühendislik Bilimleri ve Tasarım Dergisi is the property of Journal of Engineering Sciences & Design and its content may not be copied or emailed to multiple sites or posted to a listserv without the copyright holder's express written permission. However, users may print, download, or email articles for individual use. This abstract may be abridged. No warranty is given about the accuracy of the copy. Users should refer to the original published version of the material for the full abstract. (Copyright applies to all Abstracts.)

The increasing range of faults encountered by mechanical systems has brought great challenges for conducting intelligent fault diagnosis based on insufficient samples, in recent years. To tackle the issue of unbalanced samples, an improved methodology based on a generative adversarial network that uses sample generation and classification is proposed. First, 1D vibration signals are transformed into 2D images considering the features of the vibrating signals. Next, the optimized generation adversarial network is constructed for adversarial training to synthesize diverse fake 2D images according to actual sample characteristics with the generative model as a generator and the discriminative model as a discriminator. Our model uses an attenuated learning rate with a cross-iteration batch normalization layer to enhance the validity of the generator. Last, the discriminative model as a classifier is used to identify the fault states. The experimental results demonstrate that the proposed strategy efficiently improves fault identification accuracy in the two cases of sample imbalance. [ABSTRACT FROM AUTHOR]

Copyright of Ordnance Industry Automation is the property of Editorial Board for Ordnance Industry Automation and its content may not be copied or emailed to multiple sites or posted to a listserv without the copyright holder's express written permission. However, users may print, download, or email articles for individual use. This abstract may be abridged. No warranty is given about the accuracy of the copy. Users should refer to the original published version of the material for the full abstract. (Copyright applies to all Abstracts.)

Any bearing faults are a leading cause of motor damage and bring economic losses. Fast and accurate identification of bearing faults is valuable for preventing damaging the whole equipment and continuously running industrial processes without interruption. Vibration signals from a running motor can be utilized to diagnose a bearing health condition. This study proposes a detection method for bearing faults based on two types of neural networks from motor vibration data. The proposed method uses an autoencoder neural network for constructing a new motor vibration feature and a feed-forward neural network for the final detection. The constructed signal feature enhances the prediction performance by focusing more on a fault type that is difficult to detect. We conducted experiments on the CWRU bearing datasets. The experimental study shows that the proposed method improves the performance of the feed-forward neural network and outperforms the other machine learning algorithms. [ABSTRACT FROM AUTHOR]

Surface roughness is one of the most critical attributes of machined components, especially those used in high-performance systems. Online surface roughness monitoring offers advancements comparable to post-process inspection methods, reducing inspection time and costs and concurrently reducing the likelihood of defects. Currently, online monitoring approaches for surface roughness are constrained by several limitations, including the reliance on handcrafted feature extraction, which necessitates the involvement of human experts and entails time-consuming processes. Moreover, the prediction models trained under one set of cutting conditions exhibit poor performance when applied to different experimental settings. To address these challenges, this work presents a novel deep-learning-assisted online surface roughness monitoring method for ultraprecision fly cutting of copper workpieces under different cutting conditions. Tooltip acceleration signals were acquired during each cutting experiment to develop two datasets, and no handcrafted features were extracted. Five deep learning models were developed and evaluated using standard performance metrics. A convolutional neural network stacked on a long short-term memory network outperformed all other network models, yielding exceptional results, including a mean absolute percentage error as low as 1.51% and anR2 value of 96.6%. Furthermore, the robustness of the proposed model was assessed via a validation cohort analysis using experimental data obtained using cutting parameters different from those previously employed. The performance of the model remained consistent and commendable under varied conditions, asserting its applicability in real-world scenarios. [ABSTRACT FROM AUTHOR]

This study aims to develop a fault detection system designed specifically for wind turbine gearboxes. It proposes a hybrid fault diagnosis algorithm that combines scatter plot analysis with the visual geometric group (VGG) technique to identify various fault types, including gear rust, chipping, wear, and aging. To capture vibration signals, a three-axis vibration sensor was integrated with a NI-9234 DAQ card. Digital signal processing techniques were employed to actively filter out noise from the captured signals. Gaussian white noise was incorporated into the training data to enhance the noise resistance of the network model, which was then utilized for scatter plot generation. The VGG technique was subsequently applied to identify faults. The testing data were collected at two different speeds, with 1500 samples taken at each speed, totaling 3000 samples. For both training and testing, 400 samples of each fault type were employed for training, while 200 samples were allocated for testing. The test results demonstrated an overall identification accuracy of 97.7% for both the no-fault gearbox and the four-fault states, underscoring the effectiveness of the proposed methodology. [ABSTRACT FROM AUTHOR]

Bleachers play a crucial role in practical engineering applications, and any damage incurred during their operation poses a significant threat to the safety of both life and property. Consequently, it becomes imperative to conduct damage diagnosis and health monitoring of bleachers. The intricate structure of bleachers, the varied types of potential damage, and the presence of similar vibration data in adjacent locations make it challenging to achieve satisfactory diagnosis accuracy through traditional time-frequency analysis methods. Furthermore, field environmental noise can adversely impact the accuracy of bleacher damage diagnosis. To enhance the accuracy and antinoise capabilities of bleacher damage diagnosis, this paper proposes improvements to the existing Convolutional Neural Network with Training Interference (TICNN). The result is an advanced Convolutional Neural Network model with superior accuracy and robust anti-noise capabilities, referred to as Enhanced TICNN (ETICNN). ETICNN autonomously extracts optimal damage-sensitive features from the original vibration data. To validate the superiority of the proposed ETICNN, experiments are conducted using the bleacher model from Qatar University as the subject. Comparative studies under identical experimental conditions involve TICNN, Deep Convolutional Neural Networks with wide first-layer kernels (WDCNN), and One-Dimensional Convolutional Neural Network (1DCNN). The experimental findings demonstrate that the ETICNN model achieves the highest accuracy, approximately 99%, and exhibits robust classification abilities in both Phases I and II of the damage diagnosis experiments. Simultaneously, the ETICNN model demonstrates strong anti-noise capabilities, outperforming TICNN by 3% to 4% and surpassing other models in performance. [ABSTRACT FROM AUTHOR]

Monitoring the building blast vibration signal is an efficient way to determine the power of blast vibration hazards. Due to the harsh measurement environment, noise is inevitably introduced into the recorded signals. This research presents a denoising approach based on Improved complete ensemble empirical mode decomposition with adaptive noise(ICEEMDAN) and Composite Multiscale Permutation Entropy (CMPE). First, the noisy blast vibration signal is decomposed into different intrinsic mode functions using ICEEMDAN; then multiple intrinsic mode functions (IMFs) are separated into pure and noisy using CMPE, the noisy IMFs are denoised using wavelet thresholding; finally the blast wave is reconstructed using the pure and denoised mixed IMFs. The proposed approach was compared with four other approaches (CEEMDAN-CMPE, VMD-CMPE, SVMD-CMPE, and WST). The results indicate that the proposed approach has better performance and can be considered as an effective denoising method for building blast vibration signals.

Rotor-bearing systems are important components of rotating machinery and transmission systems, and imbalance and misalignment are inevitable in such systems. At present, the main challenges faced by state-of-the-art fault diagnosis methods involve the extraction of fault features under strong background noise and the classification of different fault modes. In this paper, a fault diagnosis method based on an improved deep residual shrinkage network (IDRSN) is proposed with the aim of achieving end-to-end fault diagnosis of a rotor-bearing system. First, a method called wavelet threshold denoising and variational mode decomposition (WTD-VMD) is proposed, which can process original noisy signals into intrinsic mode functions (IMFs) with a salient feature. These one-dimensional IMFs are then transformed into two-dimensional images using a Gramian angular field (GAF) to give datasets for the deep residual shrinkage network (DRSN), which can achieve high levels of accuracy under strong background noise. Finally, a comprehensive test platform for a rotor-bearing system is built to verify the effectiveness of the proposed method in the field. The true test accuracy of the model at a 95% confidence interval is found to range from 84.09% to 86.51%. The proposed model exhibits good robustness when dealing with noisy samples and gives the best classification results for fault diagnosis under misalignment, with a test accuracy of 100%. It also achieves a higher testing accuracy compared to fault diagnosis methods based on convolutional neural networks and deep residual networks without improvement. In summary, IDRSN has significant value for deep learning engineering applications involving the fault diagnosis of rotor-bearing systems.

This paper proposes a rolling bearing fault diagnosis method based on the enhanced channel attention network. The vibration signals are collected via a wireless sensor node, which are input to the neural network, and channel attention block is used to strengthen the assignment of important features, so that the attention of network is paid onto the critical fault information. Furthermore, channel attention block and residual convolution block are combined to form an enhanced channel attention network to extract the detail features. Experiment results show that the model can achieve 100% recognition accuracy for rolling bearings in various working conditions. It is also shown that the proposed new learning algorithm can provide a higher diagnosis accuracy than those state-of-the-arts in a strong disturbance environment, which reflects better robustness and generalisation ability. [ABSTRACT FROM AUTHOR]

Variational Mode Decomposition (VMD) is a useful tool for decomposing complex multi-component signals. However, one major drawback of VMD is the need to accurately determine the value of sub-signals (IMFs) before starting the process of segmentation. In fact, achieving optimal reconstruction of the denoised original signals depends on the determining optimal number of IMFs (K). This requirement poses a challenge in the capability of analyzing non-stationary or noisy signals. In this paper, a new approach to optimize the variational mode decomposition technique is proposed. This approach automatically estimates the optimal K and also effectively detects the characteristic frequencies associated with faulty bearings. This method is a combination of two algorithms which are based on cross-correlation and root mean square (RMS) statistical analysis. To confirm the efficacy of the proposed method, the bearing vibration dataset from the Case School of Engineering are used. Then, the K obtained through the proposed method are compared with other methods. The results demonstrate that the proposed approach exhibits superior robustness and precision when autonomously evaluating the optimal K for effective identification of bearing fault. [ABSTRACT FROM AUTHOR]

It has always been an important issue to diagnose mechanical equipment faults with limited training data. Specifically for the problem of bearing fault diagnosis, a multi-subspace self-attention siamese network (MSSASN) is designed for fault diagnosis with limited training data. In MSSASN, multi-subspace self-attention block is developed to assign higher weights to the fault-related information during learning. Particularly, input features are divided into multiple sub-paths in the channel dimension, and the spatial attention features are calculated separately on sub-path and then merged. In this way, the cross-channel information can be effectively learned, while multi-scale feature learning is carried out. Finally, contrastive learning is carried out on the fault features of different samples using siamese networks to deal with the problem of limited training samples. The proposed method is verified by the vibration dataset collected from the three-phase asynchronous motor experiment platform in Zhejiang University of Technology. The results show that the proposed method can identify rolling bearing faults more accurately with limited training data. [ABSTRACT FROM AUTHOR]

To address the challenge of distinguishing the health status of bearings, in this paper, a health index (HI) is developed through utilization of the multiple target time-varying black widow optimization–bidirectional gating recurrent unit (MTBWO-BiGRU) model and the Bray–Curtis distance. This index offers a visual representation of the health status of bearings, enabling more intuitive monitoring and prediction. The first step involves utilizing L1 regularization to extract effective features as degradation elements from the current bearing vibration data. Additionally, the characteristics of the initial time window of the vibration data serve as the health features. Next, the HI of the bearing is constructed by computing the Bray–Curtis distance between the bearing's degradation characteristics and health features. The cloud monitoring platform constantly tracks the health of the bearing and employs the MTBWO-BiGRU model to anticipate the forthcoming state of health. The platform generates an immediate alert when the HI of the bearing overtakes the alteration rate threshold and foresees the condition of the bearing. We compare the MTBWO-BiGRU model with the bidirectional long short-term memory (BiLSTM) and BiGRU models. The results indicate an accuracy level of 92.57%, which is evidently higher than that obtained when using the other two models. Moreover, the MTBWO-BiGRU model is lighter, demonstrating the practicality of the proposed approach. [ABSTRACT FROM AUTHOR]

Among unmanned surface vehicle (USV) components, underwater thrusters are pivotal in their mission execution integrity. Yet, these thrusters directly interact with marine environments, making them perpetually susceptible to malfunctions. To diagnose thruster faults, a non-invasive and cost-effective vibration-based methodology that does not require altering existing systems is employed. However, the vibration data collected within the hull is influenced by propeller-fluid interactions, hull damping, and structural resonant frequencies, resulting in noise and unpredictability. Furthermore, to differentiate faults not only at fixed rotational speeds but also over the entire range of a thruster's rotational speeds, traditional frequency analysis based on the Fourier transform cannot be utilized. Hence, Continuous Wavelet Transform (CWT), known for attributions encapsulating physical characteristics in both time-frequency domain nuances, was applied to address these complications and transform vibration data into a scalogram. CWT results are diagnosed using a Vision Transformer (ViT) classifier known for its global context awareness in image processing. The effectiveness of this diagnosis approach was verified through experiments using a USV designed for field experiments. Seven cases with different fault types and severity were diagnosed and yielded average accuracy of 0.9855 and 0.9908 at different vibration points, respectively. [ABSTRACT FROM AUTHOR]

Cavitation-induced vibration presents a significant challenge in vortex pumps, leading to potential damage to hydraulic components and adverse effects on pump performance. This study aims to investigate the long-term implications of such phenomena. To capture the vibration signals caused by cavitation, we utilized vibration acceleration sensors on the vortex pump and collected data at five predetermined measuring points under three different operating conditions. The analysis used two prominent techniques, fast Fourier transform (FFT) and adaptive optimal kernel time-frequency representation (AOK-TFR), to explore the frequency-domain and time-frequency characteristics of the vibration signals. The findings reveal a notable increase in frequency amplitude at each monitoring point as the flow rate rises. Under cavitation conditions, pronounced vibration characteristics are observed along the y-axis and z-axis of the volute, with maximum vibration intensities of 1.83 m/s² and 1.80 m/s², respectively. The frequency amplitude exhibits non-constant behavior in the time series. Moreover, variations in the time-frequency characteristics are identified with changing flow rates. A distinct signal with a frequency of 750 Hz manifests in the x-axis and y-axis of the volute when the head experiences a 3% reduction from the cavitation level. [ABSTRACT FROM AUTHOR]

Copyright of Journal of Fuzhou University is the property of Journal of Fuzhou University, Editorial Department and its content may not be copied or emailed to multiple sites or posted to a listserv without the copyright holder's express written permission. However, users may print, download, or email articles for individual use. This abstract may be abridged. No warranty is given about the accuracy of the copy. Users should refer to the original published version of the material for the full abstract. (Copyright applies to all Abstracts.)