Mechanical Fault Research Articles

Fractal Characteristics of the Spatial Distribution of Mine Earthquake Sources in the Vicinity of a Fault: A Case Study in the Ashele Copper Mine

Potential faults are common sensitive geological bodies that affect the safe mining of underground mines, often leading to major accidents such as rock instability and rockburst during mining. The failure mechanism of faults has been widely studied. However, due to the spatiotemporal specificity of fault occurrence, there are few theoretical and mathematical methods suitable for effective analysis in mine safety risk management. This study aims to introduce fractal theory to characterize the spatiotemporal activity fractal characteristics of induced faults intersecting the mining site and roadway during the mining process of the Ashele copper mine in China. Using microseismic systems and fractal theory, a spatiotemporal fractal model of the fault slip process is constructed, and a fractal analysis method is proposed. The fractal dimension value is calculated based on the spatiotemporal parameters of different segments and stages. The fractal dimension is used to characterize and analyze the evolution of the fault. The physical formation process of potential faults and the relationship between fractal dimension values and multiple parameters, including spatial clustering, regional distribution characteristics, and energy-release characteristics, were analyzed based on the division of events into different time stages. Discovering fractal dimension’s temporal and spatial–temporal characteristics can provide technical references for mine disaster prevention.

Prototype-assisted multiscale graph representation learning-based mechanical fault detection method under complex operating conditions

Effective anomaly detection and timely fault warning are essential to ensure the continuous and safe operation of mechanical equipment and to prevent equipment deterioration. In the unsupervised modeling and detection scenario, fault detection methods based on the autoencoder framework have been widely concerned and applied. Unfortunately, such methods can only be applied to specific or constant operating conditions, and their detection performance is greatly reduced due to the different data distribution in the face of complex operating conditions. Aiming at the problem of unsupervised fault detection under complex operating conditions, this article proposes a prototype-assisted multiscale graph representation learning-based mechanical fault detection method. First, the vibration data of the equipment is fed into the multiscale decomposition module (MDM) to obtain multiscale feature maps that can express rich detail information. Then, the multiscale feature maps are fed into the graph representation learning module (GRLM) to fully learn the potential relationships and interactions between different scales and provide a more comprehensive representation of the dynamic characteristics of the equipment. Finally, multiple MDMs and GRLMs are cascaded to construct a feature extractor to map the data of each operating condition to the latent space, and the proposed prototype-assisted strategy is used to determine the real-time state of the equipment. Case studies have been carried out on two different pieces of mechanical equipment. The experimental results show that the average accuracy of the proposed method is as high as 98.44% and 98.90%, respectively, and it maintains a low missed detection rate and zero false alarm rate in the two validation processes, which is more in line with the needs of engineering applications than other comparison methods.

Research on fault component extraction and fault type identification of rotating machinery based on MDSM and a novel convolutional neural network

Abstract Given the complexity and difficulty in extracting and recognizing multi-axis mechanical fault components, a method for fault extraction and identification based on the Multi-Axis Displacement Superposition Method (MDSM) and a Novel Convolutional Neural Network (NCNN) is proposed. In the proposed MDSM method, first, correlation analysis is used to determine the operational status of the mechanical system and to identify the location of faults in the multi-axis rotating mechanical system. Secondly, a simplified initial point selection process is introduced to segment the collected fault component. Subsequently, a signal superposition method with position offset correction is employed to perform position correction and superposition operations on the segmented signals, enhancing the accuracy of the fault signal. Finally, the front end of the superimposed signals is extracted as the fault component, completing the separation and extraction of the fault components. For the extracted fault signals, an NCNN is designed for fault-type identification. NCNN improves computational efficiency and effectively completes fault feature identification through a lightweight network architecture and a nonlinear learning rate scheduling strategy. The results of the experiment show that the proposed method can accurately determine the fault occurrence location, extract the fault components, and achieve high-accuracy fault type identification.

Southern California basin and non-basin classification algorithm for ground-motion site amplification model applications

In ground-motion modeling, the estimated level of ground shaking at any given location for an expected earthquake scenario depends on the contributions from the source component (type of fault mechanism and size of the fault slip), the path component (distance between the source and site of interest, and the geologic characteristics of that region), and the site component (the local geology at the site of interest). Each component captures some level of variability and uncertainty in the overall ground-motion estimate. In particular, the site component represents the potential amplification (or de-amplification) of the seismic waves that may lead to magnified and prolonged ground shaking at any given location. This feature is referred to as site effects and in current ground-motion models (GMMs) is dependent on the time-averaged shear wave velocity in the upper 30 m of the earth’s crust ( Vs30) and the depth to a particular shear wave velocity iso-surface (“basin depth,” zx). The latter is responsible for determining the contributions of basin effects, which is additional ground-motion amplification due to three-dimensional effects such as trapped seismic waves that lead to surface wave generation. However, an evaluation of the relationship between zx and basin locations reveals cases of misclassification that is a result of geologic variability (i.e. zx is not sufficient in differentiating basins from non-basins). The study performed in this article proposes a resolution in the form of a statistical classification model that determines the probability of a location residing within or outside a basin based on simple geologic features such as ground surface texture.

Bearing faults diagnostic in permanent magnets synchronous motor

With the widespread use of MSAPs in fields such as robotics, automotive, aerospace, medical equipment, renewable energy, etc., the preventive diagnosis of faults affecting these types of motors has become essential to avoid unplanned shutdowns that could lead to considerable financial losses. Even though the permanent magnet synchronous motor is well known for its robustness and efficiency, however it suffers from electrical or mechanical faults. Spectral analysis has become a very important tool for the detection of these faults and we will use this method to determine the frequencies characterizing the fault in MSAPs. It is important to note that the exact distribution of failures can vary depending on the specific motor and its operating conditions. The characteristic frequencies of the bearing failure are a function of the geometric parameters of the bearing and the motor rotation frequency. Unlike the induction motor, the rotation frequency of the PMSM is fixe, which makes it easier to locate the characteristic frequencies of the bearing fault. This work presents the application of spectral analysis for the diagnosis of the bearing defect and more specifically, that of the outer ring. The experimental results obtained show the interest and effectiveness of spectral analysis for the detection of bearing defects.

Application of Thermography and Convolutional Neural Network to Diagnose Mechanical Faults in Induction Motors and Gearbox Wear

Nowadays, induction motors and gearboxes play an important role in the industry due to the fact that they are indispensable tools that allow a large number of machines to operate. In this research, a diagnosis method is proposed for the detection of different faults in an electromechanical system through infrared thermography and a convolutional neural network (CNN). During the experiment, we tested different conditions in the motor and the gearbox. The induction motor was operated in four conditions, in a healthy state, with one broken bar, a damaged bearing, and misalignment, while the gearbox was operated in three conditions with healthy gears, 50% wear, and 75% wear. The motor failures and gear wear were induced by different machining operations. Data augmentation was then performed using basic transformations such as mirror image and brightness variation. Ablation tests were also carried out, and a convolutional neural network with a basic architecture was proposed; the performance indicators show a precision of 98.53%, accuracy of 98.54%, recall of 98.65%, and F1-Score of 98.55%. The system obtained confirms that through the use of infrared thermography and deep learning, it is possible to identify faults at different points of an electromechanical system.

Defect Detection Using Thermography Camera Techniques: A review

Individuals across different industries, including but not limited to agriculture, drones, pharmaceuticals and manufacturing, are increasingly using thermal cameras to achieve various safety and security goals. This widespread adoption is made possible by advancements in thermal imaging sensor technology. The current literature provides an in-depth exploration of thermography camera applications for detecting faults in sectors such as fire protection, manufacturing, aerospace, automotive, non-destructive testing and structural material industries. The current discussion builds on previous studies, emphasising the effectiveness of thermography cameras in distinguishing undetectable defects by the human eye. Various methods for defect detection, including temperature analysis and image processing algorithms, are thoroughly presented. The factors contributing to the effectiveness of thermography cameras are explored, along with their advantages over traditional inspection methods. The literature review highlights the diverse applications of thermography cameras in fault detection. The review highlights the remarkable transformation brought by thermal camera technology in mechanical system fault detection, leading to improved maintenance practices. These cameras can detect unseen irregularities, enable non-invasive testing and support hands-on system maintenance, making them indispensable tools for ensuring mechanical systems operate efficiently, reliably and safely. With the continuous advancement of technology, the integration of Industry 4.0 and IoT technologies will further enhance the capabilities of thermal cameras, ensuring elevated performance across different domains. In electrical systems, thermal cameras allow for the early identification of faults, enabling proactive maintenance to mitigate risks. Additionally, by assessing structural integrity, thermal cameras can detect thermal and insulation inefficiencies, leading to improved energy efficiency.

AI-Driven Thermography-based Fault Diagnosis in Single-Phase Induction Motor

AI-Driven Thermography-based Fault Diagnosis in Single-Phase Induction Motor

A novel unsupervised graph wavelet autoencoder for mechanical system fault detection

A novel unsupervised graph wavelet autoencoder for mechanical system fault detection

The fast bearing diagnosis based on adaptive GSR of fault feature amplification in scale-transformed fractional oscillator

From the noise-assisted perspective of stochastic resonance (SR), fractional system has been adopted to enhance the diagnostic performance of mechanical faults by utilizing the previous state information in mechanical degradation process, but the computation is extremely time-consuming. To address this challenge, we develop a fast diagnosis method leveraging the mechanism of generalized SR (GSR)-based active energy conversion in fluctuating-damping fractional oscillator (FDFO). Through the analysis of system stationary response, we propose a theoretical index known as fault feature amplification (FFA), which effectively replaces the time-consuming numerical solution in multi-parameter optimization, leading to a remarkable reduction in the time complexity of the adaptive diagnosis algorithm. This improvement brings about significant benefits, notably simplifying the diagnosis flow. Based on the results of performance evaluation in diagnosing simulated bearing signals, the proposed method exhibits a comprehensive superiority in identifying ability and diagnosis efficiency. Finally, this method has been further validated in experimental diagnosis, especially for some challenging cases, providing strong support for engineering applications, particularly in the fast diagnosis of complex operating environments.

ВІРТУАЛЬНА МОДЕЛЬ БЕЗКОНТАКТНОГО ГЕНЕРАТОРА ЗМІННОГО СТРУМУ

Currently, the main sources of alternating current electrical energy on airplanes and helicopters are contactless synchronous generators of the ГТ type with an alternating current exciter and rotating rectifiers. When conducting research on power generation systems, for example, in order to implement a diagnostic method based on monitoring electrical parameters with subsequent special analysis of the received signal, it is necessary to have reference parameters (voltages, currents, transient characteristics, voltage and current spectra) of the machine. These parameters can be obtained experimentally or by modeling, because the machine parameters in the event of typical failures cannot always be obtained experimentally. This circumstance is associated with the complexity or impossibility of physically simulating failures. For this purpose, various types of application software packages are used (for example, Matlab/Simulink, LabVIEW/Control Design and Simulation), which allow modeling complex dynamic systems, including those described by nonlinear differential equations. For example, synchronous generator models from the SimPowerSystems section of the Matlab/Simulink environment allow for good results in modeling normal generator operation. At the same time, standard models have limited failure simulation capabilities, allowing only open circuits and short circuits of windings to be simulated. To simulate other electrical and mechanical faults, standard models are usually upgraded or special ones are developed to simulate typical failures, in particular: stator winding or insulation failure, rotor winding or insulation failure, bearing failure, rotor mechanical integrity failure, stator mechanical integrity failure. As a review of publications has shown, no attention is paid to failures of the electronic part of cascade generators. The proposed work presents a virtual model of a synchronous generator, built in the MatLab/Simulink simulation environment. The model is built taking into account the design features of cascade synchronous generators of the ГТ series: the presence of three generators in the machine - a magnetoelectric exciter, an electromagnetic exciter, an electromagnetic main generator; an uncontrolled three-phase rotating rectifier; a phase-pulse voltage regulator with a controlled rectifier. The proposed model allows you to study the operation of the generator in normal operation and in the event of typical generator failures, including electronic elements. This model can be useful in studying the processes occurring in AC generation systems with ГТ series generators both in normal operation and in the event of failures. Therefore, the development of this virtual model in the Matlab/Simulink environment of a contactless synchronous generator allows you to obtain both reference parameters and parameters in the event of failures

SCML-GNN: A Graph Neural Network Model Leveraging Sensor Causality and Meta-Learning for Mechanical Fault Classification

Fault classification of mechanical equipment is a vital issue in modern industrial production. Mechanical equipment typically relies on multiple sensors to collect the operational data, which is represented as multivariate time-series data. However, existing methods for analyzing mechanical faults often overlook the causal relationships between sensors and struggle with the scarcity of labeled training samples. To address these challenges, we propose a graph neural network model leveraging sensor causality and meta-learning for mechanical fault classification (SCML-GNN). Specifically, we use transfer entropy to represent multivariate time-series data as a graph, with each sensor as a node and their causal relationships as edges. We then extract the node features using temporal convolutional layers and apply a graph neural network to learn the low-dimensional features. Additionally, graph pooling methods are used to obtain global embeddings. To further tackle the issue of limited labeled training samples, we introduce a metric-based class prototype attention mechanism within SCML-GNN. Extensive experiments conducted on three real-world mechanical equipment datasets demonstrate the superior effectiveness and efficiency of SCML-GNN in mechanical fault classification compared to the other existing methods.

Time–frequency representation of adjacent multi-component signals driven by IVKF-MPRTF and its application for mechanical equipment fault diagnosis

Rotating machinery is characterized by time-varying operating conditions and complex structures. The balance of its internal structure gradually changes with the variation of external working conditions. The vibration signals of rolling bearings that are key components exhibit obvious nonstationary characteristics at low speeds and heavy loads. Furthermore, mixed faults and modulation phenomena will generate time-varying multicomponent signals with adjacent fault characteristic frequencies. The existing time–frequency representation (TFR) algorithms for processing nonstationary signals with adjacent frequencies have limitations such as low time–frequency resolution and excessive parameter dependence. Therefore, a novel time–frequency analysis (TFA) framework for adjacent multicomponent signals driven by improved Vold–Kalman filtering-modified parameterized resampling time–frequency (IVKF-MPRTF) transform is presented in this study. First, this article introduces multicomponent signal time–frequency curve extraction to obtain the initial frequency information of the measured vibration signal, which addresses the issue of Vold–Kalman filtration (VKF) relying on a priori frequency information, and thus forms the improved Vold–Kalman filtering (IVKF). Then, the vibration signal can be precisely decomposed into adjacent components by the presented IVKF. Second, this article introduces local maximum frequency allocation operator and synchroextracting operator to compensate for the blur time–frequency resolution by parameterized resampling time–frequency transform, thus forming a modified parameterized resampling time–frequency (MPRTF) transform. By MPRTF, high-resolution TFR of multiple sets of single components can be obtained, and thus an enhanced TFR of nonstationary multicomponent signals is achieved through time–frequency fusion. Finally, the superiority of the adjacent multicomponent signals identification by the proposed IVKF-MPRTF is verified through the experimental signal analysis.

Fault diagnosis of HVCB via the subtraction average based optimizer algorithm optimized multi channel CNN-SABO-SVM network

The mechanical fault diagnosis of HVCB is important to ensure the stability of electric power systems. Aiming at the problem of poor diagnostic performance of deep learning methods under limited samples, this paper proposes an HVCB operating mechanism fault diagnosis model (multi-channel CNN-SABO-SVM, MCCSS) based on multimodal data fusion features and Subtraction-Average-Based Optimizer (SABO). This model extracts and fuses features from the input two-dimensional data using a multi-channel CNN network and then uses the multimodal data fusion features to diagnose HVCB faults. Additionally, the SVM is used instead of the Softmax classifier to classify the fused features of vibration and sound, compensating for the poor diagnostic performance and generalization ability of the CNN network in small sample data scenarios. To further enhance the fault diagnosis performance of the SVM, the SABO is introduced for hyperparameter optimization of the SVM classifier. An HVCB fault test platform was established to train and test the model with limited data. The experimental results show that, compared with the multi-channel CNN-SVM and the CNN model based on unimodal signals, the proposed multi-channel CNN-SABO-SVM model improves the accuracy by 2.66% and 10.66%, respectively, and effectively addresses the challenge of circuit breaker fault diagnosis with limited samples.

Few-shot sample multi-class incremental fault diagnosis for gearbox based on convolutional-attention fusion network

Few-shot sample multi-class incremental fault diagnosis for gearbox based on convolutional-attention fusion network

A three-stage cross-domain intelligent fault diagnosis method for multiple new faults

Abstract Recently, transfer learning (TL) has been widely investigated to tackle the cross-domain fault diagnosis issue in machinery, and most research works follow the same assumption that the diagnosis domain shares the same fault categories. However, due to the randomness and complexity of mechanical faults, the new fault modes usually occur unexpectedly in the actual scenarios. The emergence of new faults also presents severe challenges to TL. In response to these challenges, a three-stage cross-domain intelligent fault diagnosis method is presented in this article. First, partial domain alignment is achieved based on an improved target weighted mechanism, and an outlier identifier is constructed to automatically separate the new fault classes. Then, an unsupervised learning model with silhouette coefficients is built to determine the number of new fault categories. Finally, the simulation signals are further adopted to distinguish the specific fault categories. Sufficient experiments on axial piston pump and public bearing datasets validate that the proposed method can predict a number of new fault categories and identify specific fault categories. The results indicate that the proposed method outperforms the other methods and has promising practical applications in fault diagnosis with multiple new faults.

Aerospace Equipment Fault Diagnosis Method Based on Fuzzy Fault Tree Analysis and Interpretable Interval Belief Rule Base

The stable operation of aerospace equipment is important for space safety, and the fault diagnosis of aerospace equipment is of practical significance. A fault diagnosis system needs to establish clear causal relationships and provide interpretable determination results. Fuzzy fault tree analysis (FFTA) is a flexible and powerful fault diagnosis method, which can deeply understand causes and fault mechanisms. The interval belief rule base (IBRB) can describe uncertainty. In this paper, an interpretable fault diagnosis model (FFDI) for aerospace equipment based on FFTA and the IBRB is presented for the first time. Firstly, the initial FFDI is constructed with the assistance of FFTA. Second, a model inference is implemented based on an evidential reasoning (ER) parsing algorithm. Then, a projection covariance matrix adaptive evolutionary strategy algorithm with an interpretability constraints (IP-CMA-ES) optimization algorithm is used for optimization. Finally, the effectiveness of the FFDI is verified by a flywheel dataset. This method ensures the completeness of the rule base and the interpretability of the model, avoids the problem of exploding certain combinations of rules, and is suitable for the fault diagnosis of aerospace equipment.

Vibration monitoring of rotating shafts using DIC and compressed sensing

Vibration monitoring of rotating shafts using DIC and compressed sensing

A dual fiber grating acceleration sensor based on elastic ring structure of “8” shape

A dual fiber grating acceleration sensor based on elastic ring structure of “8” shape

RISC Fault of Non‐Salient Wound‐Rotor Synchronous Generator: A Detection Method and Effect Analysis of Insulation

Rotor interturn short circuit (RISC) fault is one of the most prevalent fault types in the non‐salient wound‐rotor synchronous generator (WRSG). In the paper, the theoretical mechanism of the RISC fault is studied in detail. Building upon this understanding, from the fault detection and harm consequences, the further work of this article includes proposing a RISC detection method and analyzing the influence of RISC on the insulation thermal damage. On the one hand, a new detection for RISC fault is first proposed based on the phase voltage. The proposed detection method innovatively allows for accurate determination of both the severity and location of RISC faults. On the other hand, the loss and the temperature of the stator core‐winding system are comprehensively investigated with varied RISC degrees and positions. The thermal responses of the winding insulation are obtained under the joint action of the external magnetic heat source and the internal electric heat source. The study identifies critical points of insulation vulnerability, particularly noting areas prone to noise and connection joints. This research contributes to the timely online identification of RISC faults and supports preventive maintenance strategies for protecting winding insulation. © 2024 Institute of Electrical Engineers of Japan and Wiley Periodicals LLC.