Fault Diagnosis Research Articles

Convolutional neural networks in fault diagnosis of industrial internet of things

Convolutional neural networks in fault diagnosis of industrial internet of things

Pre-connected and trainable adjacency matrix-based GCN and neighbor feature approximation for industrial fault diagnosis

Industrial fault diagnosis methods based on graph convolution network (GCN) becomes a hot topic for its great feature extraction ability to multivariate time-series data. However, GCNs ignore inter-sample temporality when constructing the adjacency matrix (AM), leading to low prediction accuracy. A novel fault diagnosis method based on pre-connected and trainable AM-based GCN and neighbor feature approximation (PTGCN-FA) is proposed at the node-level task. Firstly, PTGCN-FA introduces the temporal nearest neighbors into spatial nearest neighbors to pre-connect and construct the AM. Then, the AM is trained only where the samples are connected, which makes the best weights obtained and reduces the time complexity of the model. Finally, after the GCN layers, the trained AM is introduced into the approximation of features, which are neighbors in the original sample space. Two process industry cases are carried out, and the simulation results including diagnosis accuracy, confusion matrix, study to the ratio of labeled data and an ablation experiment verify PTGCN-FA has more efficient and accurate diagnostic performance than related methods. Additionally, the analysis of the temporal neighborhood weight parameter shows that the performance of fault diagnosis can be improved by considering both temporal and spatial information between samples.

Phase-based video vibration measurement and fault feature extraction method for compound faults of rolling bearings

Vibration characterization of rotating machinery is crucial for determining rotational and failure frequencies. Traditional contact measurement methods have limitations, while high-speed cameras offer a non-contact alternative for measuring target vibrations, spatial phase-based techniques have recently been widely used in detecting subtle vibrations and show good robustness to imaging noise. In this paper, a vision-based vibration extraction method for rotating machinery is proposed, aiming at extracting minor vibrations of bearings from them and further analyzing their fault frequencies. To address the challenge of separating a single fault frequency from the extracted compound faulty vibration signals, kurtosis is employed to analyze the inverse convolution period of multiple periodic components, and the MOMEDA filter length is optimized using the golden section algorithm. MOMEDA is then used to enhance each periodic pulse separately, and fault frequency is obtained through envelope spectrum demodulation. In the experimental part, a phase-based method is used to extract minor vibration displacements from rotor vibration video, which is subsequently compared with eddy current sensors in both time and frequency domains to verify the accuracy of the proposed method in extracting vibration displacements based on vision. Finally, vibration signals are extracted from the bearing compound fault video, and the single fault frequency characteristics of the bearing are successfully separated using the adaptive MOMEDA method, which provides an efficient and reliable method in the field of rotating machinery fault diagnosis.

Fixed-time active fault-tolerant control for a class of nonlinear systems with intermittent faults and input saturation

In this paper, the problem of fixed-time active fault-tolerant control is studied for systems with sector-bounded nonlinearities, intermittent faults, and input saturation. Since intermittent faults appear and disappear randomly, a fixed-time scheme is considered in the active fault-tolerant control algorithm, composed of detection, isolation, estimation, and the control unit. Utilizing homogeneity-based observers, the fixed-time state estimation is available in the presence of unknown but bounded disturbances, and a fault diagnosis unit is proposed. An input saturation compensator is introduced to analyze the effect of input saturation, and its auxiliary variables are used in the reconfigurable control law. The fault-tolerant controller, which is constructed via the information provided by the fault diagnosis unit and saturation compensator, has two switching modes. As a consequence, intermittent faults are compensated via the designed active fault-tolerant control method and the system reaches practical stability with the entire convergence time bounded in a fixed time. Finally, the example of a heat control system is exploited to demonstrate the effectiveness of the developed active fault-tolerant control scheme.

Dynamic Temporal Denoise Neural Network with Multi-Head Attention for Fault Diagnosis Under Noise Background

Fault diagnosis plays a crucial role in maintaining the operational safety of mechanical systems. As intelligent data-driven approaches evolve, deep learning (DL) has emerged as a pivotal technique in fault diagnosis research. However, the collected vibrational signals from mechanical systems are usually corrupted by unrelated noises due to complicated transfer path modulations and component coupling. To solve the above problems, this paper proposed the dynamic temporal denoise neural network with multi-head attention (DTDNet). Firstly, this model transforms one-dimensional signals into two-dimensional tensors based on the periodic self-similarity of signals, employing multi-scale two-dimensional convolution kernels to extract signal features both within and across periods. Secondly, for the problem of lacking denoising structure in traditional convolutional neural networks, a temporal variable denoise (TVD) module with dynamic nonlinear processing is proposed to filter the noises. Lastly, a multi-head attention fusion (MAF) module is used to weight the denoted features of signals with different periods. Evaluation on two datasets, Case Western Reserve University bearing dataset (single sensor) and Real aircraft sensor dataset (multiple sensors), demonstrates that the DTDNet can reduce the useless noises in signals and achieve a remarkable improvement in classification performance compared with the state-of-the-art method. DTDNet provides a high-performance solution for potential noise that may occur in actual fault diagnosis tasks, which has important application value.

Fault diagnosis of transformer based on XGBoost

Abstract. With the continuous development of technology, the fault diagnosis of transformer is also continuously advancing. Traditional fault diagnosis methods such as the three-ratio method and the characteristic gas method have the disadvantages of missing coding and limited identification capabilities. Data-driven fault diagnosis methods can integrate various information for data analysis to accurately diagnose the type of fault. This paper proposes a fault diagnosis method for transformers based on the XGBoost model. First, the input data is selected for the ratio feature engineering, and the obtained ratios are also used as input data, and then the input data is normalized for data preprocessing. Secondly, the XGBoost model is introduced, which minimizes the prediction errors in each round of gradient boosting, thereby improving the accuracy and recall rate. Finally, through case analysis, comparisons are made with other methods in various performance indicators, and the accuracy is improved. The results show that the XGBoost is more accurate in the fault diagnosis of transformers, proving that the research has certain value and significance.

A Bearing Fault Diagnosis Model with Convolutional Cross Transformer and ResNet18

Abstract In the industrial field, malfunction of rotating machinery, especially bearings, can cause significant economic losses to enterprises. Addressing the limitations of traditional fault diagnosis methods, such as poor generalization performance and low noise resistance, this paper introduces a fault diagnosis model that parallels the Cross Convolutional Transformer and ResNet18 (CCTAR). The proposed CCTAR utilizes two feature extraction channels, aimed at balancing the extraction of local and global features, and the specially designed convolutional cross-decoding layer has excellent noise resistance, surpassing traditional multi-layer Transformer encoding layers with a single-layer structure. CCTAR achieves commendable recognition accuracy across multiple datasets and maintains high accuracy in noisy environments. Furthermore, transfer learning experiments have demonstrated the proposed model's capability to achieve superior fault diagnosis performance across different working conditions with a limited number of samples, highlighting its practical significance.

End-to-End Intelligent Fault Diagnosis of Transmission Bearings in Electric Vehicles Based on CNN

Environmental noise and transmission components can cause significant interference in vibration signals, rendering the extraction of bearing fault features challenging in service scenarios. Traditional fault diagnosis methods rely heavily on professional domain knowledge, prior models, and signal preprocessing methods. The accuracy of fault diagnosis depends on the quality of the fault-sensitive features extracted by vibration signal preprocessing methods. An improved convolutional neural network (CNN) end-to-end intelligent fault diagnosis model based on raw vibration data (RVDCNN) is proposed. The time-domain vibration signal of the transmission bearing is converted into a continuous two-dimensional numerical matrix, and a two-dimensional CNN model is constructed through network structure optimization. The original time-domain vibration signal numerical matrix of the bearing is trained and tested to extract and learn abstract fault features of different fault types, and then the fault classification of the bearing is achieved. To verify the generalizability of the RVDCNN intelligent fault diagnosis model, it is applied to the recognition of rolling bearings in the two-speed mechanical automatic transmission of electric vehicles, achieving recognition accuracy of 99.11% for seven types of bearings.

Novel sparse auto-encoder framework with pseudo-labeled reinforcement for cross domain fault diagnosis with imbalanced samples

Abstract By applying diagnostic expertise from one area to another that is closely related, transfer fault diagnosis is an efficient strategy for guaranteeing the secure and dependable operation of mechanical equipment. However, the majority of rotating machinery monitoring data are gathered in industrial applications under normal operating settings, which leads to an imbalance between positive and negative samples. Consequently, performing high-precision fault diagnosis with imbalanced data and different working conditions becomes challenging due to the escalating difficulty and cost of acquiring labeled fault samples. For cross domain fault diagnosis, an enhanced deep transfer sparse auto-encoder (SAE) framework is provided. This approach leverages a SAE to delve deeper into fault features within imbalanced data and reconstruct the samples accordingly. Furthermore, the research proposes a feature domain penalty term to facilitate cross-domain training by aligning the distribution of the source and target domain data which can reduce data distribution deviation. A pseudo-labeled reinforcement training method is presented to further improve cross-domain classification accuracy with imbalanced samples. Extensive experiments were conducted on two datasets to assess the proposed method. The results were compared with other algorithms, demonstrating the effectiveness and superiority of the proposed approach.

Fault diagnosis of floating offshore wind turbine based on bidirectional long short-term memory networks with sliding window

Owing to its ability to harvest more power in deeper waters than other types of offshore wind turbines, the floating offshore wind turbine (FOWT) has attracted significant interest from both academics and industry. However, deterioration in system performance and unscheduled shutdowns will significantly increase operation and maintenance costs. To avoid system damage and ensure operation safety, there is a great demand for developing FOWT fault diagnosis to detect and identify faults as early as feasible. Due to the enormous complexity of FOWT including dynamics and nonlinearity, effective FOWT fault diagnosis poses a significant challenge. This paper develops a novel FOWT fault diagnosis method using bidirectional long short-term memory (BiLSTM) with sliding window. Firstly, the dynamics are embedded by stacking time-dependent measurements using the sliding window technique. Then, the augmented data is fed into a BiLSTM network to exploit the nonlinearity of FOWT. Consequently, a softmax classifier is used for fault diagnosis. A high-fidelity 10 MW spar FOWT benchmark based on the NREL Fatigue, Aerodynamics, Structures, and Turbulence (FAST) model is used to validate the capability and effectiveness of the proposed method. Experimental results show that the proposed method outperforms other related methods in diagnosing critical faults in FOWTs under various operating conditions.

AI-Augmented Fault Detection and Diagnosis in VLSI Circuits: A Step toward Intelligent Chip Design

Fault detection and diagnosis are critical challenges in Very Large Scale Integration (VLSI) circuits, where even minor defects can cause significant performance degradation or system failure. Traditional fault detection methods often struggle with scalability and real-time analysis as circuits grow increasingly complex. This paper proposes an AI-augmented framework that leverages machine learning algorithms and neural networks to enhance the fault detection process in VLSI circuits. The proposed model not only identifies and classifies faults with high accuracy but also provides root-cause diagnostics, significantly reducing testing time and maintenance costs. Simulation results demonstrate the effectiveness of the AI-based approach in comparison with conventional techniques, showcasing improved fault coverage, faster detection, and scalability for modern chip designs. This study serves as a step toward the integration of intelligent diagnostics into chip manufacturing, paving the way for more robust and self-healing electronic systems.

Condition Monitoring of Electrical Transformers Using the Internet of Things: A Systematic Literature Review

The adoption of Internet of Things (IoT) technology for transformer condition monitoring is increasingly replacing traditional methods. This systematic review aims to evaluate the existing research on IoT frameworks used in transformer condition monitoring, providing insights into their effectiveness and research trends. This review seeks to identify the leading IoT frameworks employed in transformer condition monitoring; analyze the key research objectives, methods, and outcomes; and assess the global research distribution and technological tools used in this field. A systematic literature review was conducted by searching published databases using keywords related to “Internet of Things”, “transformers”, “condition monitoring”, and “fault diagnosis”. The search spanned publications released between 2014 and 2024, yielding 262 articles. Of these, 120 met the predefined review criteria and were included for further analysis. This review found that Arduino boards are the most used microcontrollers for monitoring and analyzing transformer operational parameters, with Arduino IDE 1.8 being the predominant software for programming. The primary research focus in the reviewed literature is the identification of transformer faults. The geographical distribution of research contributions shows that India leads with 65% of the studies, followed by China (11%) and Pakistan (5%). The findings indicate a strong global interest in developing IoT-based transformer condition monitoring systems, particularly in India. This review highlights the potential of IoT technologies to enhance transformer monitoring and diagnostics. The insights gained from this review can guide future research and the development of more advanced IoT frameworks for transformer condition monitoring.

Fault Diagnosis Method for Marine Electric Propulsion Systems Based on Zero-Crossing Tacholess Order Tracking

In marine electric propulsion systems (MEPS) driven by variable-frequency drives, motor current signals often exhibit complex modulation components, ambiguous spectra, and severe noise interference, rendering it challenging to extract fault-related modulation components. To address this issue, we propose a zero-crossing tacholess order tracking method based on motor current signals. This method utilizes zero-crossing estimation of the instantaneous frequency to perform angular resampling of stator current signals and demodulates the envelope spectrum to extract fault characteristic spectra, enabling the diagnosis of mechanical faults in MEPS. Given the synchronization of the synchronous motor speed with the inverter fundamental frequency, this method estimates instantaneous frequencies in the time domain without requiring integration or time–frequency representation, which is simple and computationally efficient. Data validation on a small-scale marine electric propulsion test platform demonstrates that the proposed method exhibits good robustness under variable-speed conditions and effectively detects imbalance faults caused by propeller breakages and gear faults resulting from bevel gear tooth defects. Therefore, the proposed method can be applied to diagnose faults in downstream mechanical equipment driven by motors.

Rolling bearing fault diagnosis method based on dynamic simulated model from source domain to target domain with improved alternating transfer learning

Rolling bearing fault diagnosis method based on dynamic simulated model from source domain to target domain with improved alternating transfer learning

Power Transformer Fault Diagnosis Based on Random Forest and Improved Particle Swarm Optimization–Backpropagation–AdaBoost

This paper proposes a novel fault diagnosis methodology for oil-immersed transformers to improve the diagnostic accuracy influenced by gas components in power transformer oil. Firstly, the Random Forest (RF) algorithm is utilized to evaluate and filter the raw data features, solving the problem of determining significant features in the dataset. Secondly, a multi-strategy Improved Particle Swarm Optimization (IPSO) is applied to optimize a double-hidden layer backpropagation neural network (BPNN), which overcomes the challenge of determining hyperparameters in the model. Four enhancement strategies, including SPM chaos mapping based on opposition-based learning, adaptive weight, spiral flight search, and crisscross strategies, are introduced based on traditional Particle Swarm Optimization (PSO) to enhance the model’s optimization capabilities. Lastly, AdaBoost is integrated to fortify the resilience of the IPSO-BP network. Ablation experiments demonstrate an enhanced convergence rate and model accuracy of IPSO. Case analysis using Dissolved Gas Analysis (DGA) samples compares the proposed IPSO–BP–AdaBoost model with other swarm intelligence optimization algorithms integrated with BPNN. The experimental findings highlight the superior diagnostic accuracy and classification performance of the IPSO–BP–AdaBoost model.

Intelligent Diagnosis of Bearing Failures Based on Recurrence Quantification and Energy Difference

Bearing health is key for maintaining good performance and safety in rotating machinery. As the diagnosis of mechanical faults develops toward intelligence and automation, accurate and systematic fault diagnosis algorithms are imperative. Focusing on the diagnosis of rolling bearing failures, this study utilizes a sliding time window to extract essential data segments. A series of signal processing techniques, including filtering, amplitude–frequency analysis, Hilbert envelope analysis, and energy analysis, is applied to establish a comprehensive dataset. For extraction of the hidden properties of the data, the recurrence quantity spectrum is defined for the input of the neural network. The goal is to obtain a cleaner dataset with enhanced features. A convolution neural network is constructed. Different activation functions in the activation layer are compared for better fault diagnosis algorithms. The established feature matrices are specifically defined to accurately identify the subtlest defects of bearings, thereby facilitating early detection. The proposed procedure distinguishes various fault modes. As for the multidimensional complexities of fault signals, this study carries out a comprehensive comparison of energies, recurrence quantification, and amplitude–frequency characteristics of bearing fault detection to assess the accuracy, computational efficiency, and robustness of bearing fault diagnosis. The proposed method and bearing fault detection procedures have potential in practical applications.

Dynamic model-based intelligent fault diagnosis method for fault detection in a rod fastening rotor

A complete fault sample database is of great significance for the intelligent fault diagnosis method of rod fastening rotor. However, the lack of fault samples makes the fault diagnosis results unbelievable. To solve this issue, the dynamic model-based intelligent fault diagnosis method is established for a rod fastening rotor, and the fault sample database is enriched by numerical simulations. First, the lumped parameter model of the rod fastening rotor system is constructed and further updated using Euclidean Distance between measurement and numerical simulation of the intact system. Second, mathematical models of various fault types are incorporate into the updated model to obtain numerical simulation fault samples. Thirdly, the utilization of numerical simulation fault samples is severed as training data to the artificial intelligence (AI) models and the unknown measurement test samples will be finally classified. In this paper, Support Vector Machine, Random Forest, Bayesian Network and Decision Tree are selected as the typical AI models. Subsequently, the feasibility of classification is validated by the test bench of the rod fastening rotor system, and the problem of insufficient fault samples can be solved.

An Enhanced Fault Localization Technique for Distribution Networks Utilizing Cost-Sensitive Graph Neural Networks

Accurate and timely fault diagnosis is of great significance for the stable operation of a distribution network. Traditional artificial intelligence-based localization methods rely heavily on large-scale labeled datasets, making them prone to being affected by distributed generators. To address this issue, this study proposes a fault location method for distribution systems using a cost-sensitive graph attention network model. In this approach, the physical structure of the distribution network is considered a crucial constraint for model training, thereby enhancing its information perception capabilities. Specifically, the electrical nodes and lines of the distribution network are mapped to the vertices and edges within the graph attention network, with the attention weights determined by the correlation of adjacent node fault characteristics, thus improving the sensitivity to grounding faults. Additionally, a cost-sensitive matrix is used to balance class distribution, enhancing the robustness and generalization ability of the model. Fault localization experiments were conducted on the IEEE-33 bus distribution system to validate the effectiveness of the proposed fault localization method. Factors such as data disturbance, varying fault grounding resistances, and distributed power supply access were employed to assess the model’s anti-interference performance. The experimental results demonstrate that the fault location method exhibits high positioning accuracy and excellent robustness.

An Industrial Fault Diagnosis Method Based on Graph Attention Network

An Industrial Fault Diagnosis Method Based on Graph Attention Network

Compact and Simple Antenna Inspection Testing Range for 5G Smartphones

Abstract Over-the-air (OTA) antenna performance evaluation in the millimeter-wave (mmWave) band often relies on Compact Antenna Test Range (CATR) systems, which require large reflectors and a substantial measurement range volume. Such setups are impractical for rapid testing in space-constrained environments such as antenna fault diagnosis in smartphone manufacturing facilities. This paper presents a novel design methodology for highly compact antenna inspection testing range utilizing offset parabolic reflectors specifically designed for 5G smartphones. The proposed design system employs three small offset parabolic reflectors configured to account for beam tilting caused by smartphone back covers, eliminating the need to physically reposition the smartphone during testing. By directing electromagnetic fields toward a fixed receiver antenna from multiple angles corresponding to different antenna module positions, the system streamlines the test procedure, significantly reducing measurement time and measurement range volume. The proposed antenna inspection testing range provides a cost-effective and efficient solution for 5G antenna fault diagnosis in manufacturing facilities, ensuring rapid and reliable inspection while minimizing space requirements on the production line.