Fault Knowledge Research Articles

Intelligent Fault Diagnosis for Aero-Engine Lubrication Systems Using Knowledge Graph and Deep Learning Techniques

Due to the complex structure and function of aero-engine lubrication system, fault diagnosis based on the existing health management system is not explicable and highly dependent on experts' experience. A method for constructing aeroengine lubrication system fault knowledge graph was proposed in this paper. Based on the experts' knowledge, the concept of lubrication system fault knowledge graph ontology was designed. With the help of deep learning techniques such as BiLSTM and CRF, we achieved the automatic extraction of unstructured knowledge. Next, based on the Cosine Distance and Jaccard coefficient, multi-source heterogeneous fault knowledge fusion was realized. In the end, intelligent fault knowledge question answering and fault attribution analysis are realized based on the building of areo-engine lubrication system fault knowledge graph. The application results show that the knowledge graph technology can realize the utilization of prior knowledge of lubrication system faults and the explanation of fault causes. And the knowledge graph technology has a good application prospect in the field of intelligent fault diagnosis.

SFDA-T: A novel source-free domain adaptation method with strong generalization ability for fault diagnosis

Currently, source free domain adaptation (SFDA) methods are employed to address the issue of inaccessible source domain data (SDD) in transfer learning. However, existing SFDA methods often suffer from overfitting to specific domains, leading to poor generalization ability in the target domain. To address these challenges, this paper proposes a novel SFDA method named SFDA-T for fault diagnosis. Specifically, a Transformer-CNN-based feature extractor is constructed, to mine the transferable feature knowledge of faults in the SDD. The approach reduces the overfitting of the model to domain-specific information and improves model’s generalization ability. In addition, the feature attention loss is designed to calculate attention weights of the sample features to increase the model’s attention to the crucial feature regions in the target domain. A source similarity guided exponential loss is developed to guide target samples based on the decision boundaries of the source domain, facilitating cluster alignment of target sample categories and expanding distances between different categories. Furthermore, a self-training pseudo-labeling constraint is employed to reduce the effect of incorrect label matching and further constrain the model. The results of the experiments on gearboxes and bearings indicate that the proposed method achieves high fault diagnosis accuracy while effectively decoupling from SDD.

A research and application of a digital failure knowledge model for nuclear power plant

Abstract The accumulation and application of knowledge about equipment component failures in nuclear power plant unit systems is an important way for domestic and foreign nuclear power plants to improve unit availability and economic benefits. Based on the existing equipment component knowledge models in China and the experience of digital transformation, this paper proposes a digital fault knowledge model construction process and method for nuclear power plant components, which is of great significance for standardizing and guiding domestic nuclear power plant knowledge models.

Federated transfer learning-based distributed fault diagnosis method for rolling bearings

Abstract Current methods for bearing fault diagnosis often fall short in addressing data privacy concerns and typically rely on one-to-one transfer strategies, which are inadequate for achieving knowledge transfer in distributed environments. To address this issue, a distributed fault diagnosis method for rolling bearings based on federated transfer learning is proposed. This method ensures data privacy while integrating fault knowledge from multiple domains, thereby enabling more efficient knowledge transfer. Specifically, a Domain Adversarial Neural Network (DANN) is introduced as the base model within the federated learning framework. Additionally, Maximum Mean Discrepancy (MMD) is incorporated into the DANN to enhance the transfer of fault knowledge. Finally, a dynamic weighting parameter update method based on MMD is designed to evaluate the feature discrepancies between source and target domains, thereby updating the parameters of the federated framework and achieving global model aggregation. Experimental results on two bearing datasets demonstrate that the proposed method excels in both distribution alignment and fault diagnosis.

Reserving embedding space for new fault types: A new continual learning method for bearing fault diagnosis

In complex operating environments, rotating equipment may continually generate new fault categories, affecting the safety of equipment operation, and the number of collected fault samples is limited, which make it difficult to establish a reliable diagnostic model. Few-shot continual learning (FSCL) can continuously learn and summarize fault knowledge from limited samples. These traditional FSCL models, when learning from limited samples of new types of bearing faults, exhibit model overfitting and catastrophic forgetting issues after self-updating (new model). This phenomenon significantly limits the reliability of traditional models. This paper presents a new approach coined as reserving embedding space for new fault types (RESNFT) for continual learning of newly occurred bearing faults for machinery fault diagnosis, which includes several key and new ideas. Preliminary, a class prototype (prototype vector) is derived to replace a classifier weight. Then, embedding space reserved by forcibly assigning real fault samples to both real and virtual categories can mitigate catastrophic forgetting. Finally, virtual faults formed by sample mixing guide new fault samples to the reserved embedding space, and realize continuous classification of new faults. Results show that the RESNFT can effectively alleviate the catastrophic forgetting and overfitting problems, and it is superior to other existing methods.

Analysis of Fault Events in Rail Transit Vehicle Traction Systems Based on Knowledge Graph Reasoning

Fault text records provide detailed information on faults and handling steps, which are valuable for fault analysis. However, the different individuals’ recording styles can lead to ambiguities, and it is challenging to uncover potential fault associations in complex systems. To address these issues, this paper proposes a novel method for fault information extraction and analysis. Firstly, to tackle the problem of ambiguous boundaries between entities in fault texts, an integration algorithm is employed to accurately recognize fault entities considering contextual semantic features to establish fault knowledge graph (FKG). Then, a Relational Graph Convolutional Networks (RGCN) is improved with Graph Attention Networks (GAT) for sparse nodes caused by specific types of faults, to dynamically adjust the weight distribution of node learning, inferring potential links within the graph. The proposed method was validated using actual fault records from the traction system of rail transit vehicles, and contributes a reference for the mining and analysis of fault records in complex systems.

Early anomaly detection of wind turbine gearbox based on SLFormer neural network

Detecting faults in wind turbines at early stage is of great significance in improving the economic efficiency of wind farms. However, the widely used fault detection techniques are mostly based on traditional deep learning frameworks, which are limited in handling long-term dependencies, leading to constraints in global feature extraction. Additionally, current studies often build models based on the operational performance of an individual wind turbine, limiting the generalizability of the models. Therefore, this study introduces an improved model called SLFormer, which integrates Long Short-Term Memory into the Transformer encoder for early fault detection in the gearbox. Simultaneously, a novel fault detection strategy based on SCADA data analysis and transfer learning is proposed. Case studies from three wind farms indicate that the SLFormer model significantly outperforms six other popular prediction models in terms of stability and accuracy in modeling normal behavior and exhibits high robustness to random disturbances in SCADA data. The SLFormer model can predict gearbox anomalies twenty-one days in advance and effectively avoids false fault reports. The proposed strategy can help the model acquire fault knowledge from multiple wind farms, thus creating a framework with generalizability.

Research on Fault Detection and Automatic Diagnosis Technology of Water Hammer in Centrifugal Pump

Centrifugal pumps are important equipment in industrial production, and their safe and reliable operation is of great significance to water supply and industrial safety. During the use of centrifugal pumps, faults such as bearing damage, blade wear, shaft imbalance, shaft misalignment and water hammer often occur. Among them, although water hammer faults occur at a low frequency, they are difficult to monitor and pose significant risks to valve and pipeline interfaces. This article analyzes the causes, mechanisms and phenomena of water hammer faults in centrifugal pumps, designs a monitoring method to effectively monitor the vibration signal of the centrifugal pumps, extracts vibration characteristics to determine and record water hammer events, designs monitoring and diagnostic models for the edge layer and server side, and establishes an experimental verification testing system. The test results show that, under the conditions of simulating water hammer faults, after high-pass filtering of the collected vibration data, the kurtosis index, pulse index and margin index all exceed twice the threshold, and both sensors emit water hammer alarms. The designed data acquisition method can capture water hammer signals in a timely manner, and the analysis model can automatically identify water hammer faults based on existing fault knowledge and rules. This fully demonstrates the scientific and effective nature of the proposed centrifugal pump fault monitoring method and system, which is of great significance for ensuring the safe operation and improving the design of centrifugal pumps.

A new indirect transfer fault diagnosis method based on feature separation

The existing fault diagnosis methods based on deep transfer learning achieve domain adaptation by matching shared features between the source and target domains. However, there are different private characteristics between the source domain and the target domain, and traditional cross-domain fault diagnosis algorithms ignore the impact of private attributes in the target domain on classification tasks. Meanwhile, the existence of private features causes a negative transfer problem for domain adaptation methods in domains with large aligned domain gaps. In practical applications, indirect transfer between the source domain and target domain can reduce the limitations of inherent attributes and enhance the guidance of private features in the target domain for classification tasks, which can enrich the transfer of fault knowledge and improve the effectiveness of cross-domain fault diagnosis. Therefore, an indirect transfer fault diagnosis method based on feature separation (ITFS) is proposed. Firstly, an attention-enhanced feature separation framework is proposed, which constructs a set of joint attention convolutional autoencoders (CAEs) to separate features from two domains, avoiding the influence of domain shift in private features. Secondly, to better adapt the two domains, an indirect domain adaptation method based on Gaussian distribution is proposed, which combines Gaussian indirect transfer learning with pseudo-labeling enhancement to reduce the domain gap in the two domains. Six transfer tasks on the PU datasets demonstrate that compared to MCD, the average accuracy of the proposed method is improved by 14.09 %, and it is superior to other popular existing methods. Experiments on three publicly available datasets have shown that this method outperforms other popular existing methods.

Performance-Based Hierarchical Fault-Tolerant Control for Closed-Loop Systems With Multiplicative Faults: A Data-Driven Design Method.

Fault-tolerant control (FTC) is vital for the safety and reliability of automatic systems. Most of the existing FTC methods are developed for open-loop systems subject to additive faults, regardless of the widely present control loops and multiplicative faults within systems. In this article, a performance-based FTC strategy is proposed for the closed-loop systems with multiplicative faults. Considering the high efforts in modeling complex systems, the proposed FTC strategy is realized in the data-driven context. Specifically, a nominal feedback-feedforward controller is first established for the fault-free systems. By selecting the system stability and reference tracking behavior as the key performance indices, two performance evaluators are constructed to detect and classify the occurred multiplicative faults based on the fault-induced effects on the system performance. Then, with the aid of the coprime factorization technique, the multiplicative faults, in the form of additive perturbations to the system coprime factors, are estimated utilizing the closed-loop process data. Furthermore, based on the fault knowledge, a hierarchical fault-tolerant tracking controller is developed according to the levels of system performance degradations, where the functional controller parameters are reconfigured with different priorities. Finally, case studies are provided to validate the effectiveness of the proposed method.

A related degree-based frequent pattern mining algorithm for railway fault data

It is of great significance to improve the efficiency of railway production and operation by realizing the fault knowledge association through the efficient data mining algorithm. However, high utility quantitative frequent pattern mining algorithms in the field of data mining still suffer from the problems of low time-memory performance and are not easy to scale up. In the context of such needs, we propose a related degree-based frequent pattern mining algorithm, named Related High Utility Quantitative Item set Mining (RHUQI-Miner), to enable the effective mining of railway fault data. The algorithm constructs the item-related degree structure of fault data and gives a pruning optimization strategy to find frequent patterns with higher related degrees, reducing redundancy and invalid frequent patterns. Subsequently, it uses the fixed pattern length strategy to modify the utility information of the item in the mining process so that the algorithm can control the length of the output frequent pattern according to the actual data situation and further improve the performance and practicability of the algorithm. The experimental results on the real fault dataset show that RHUQI-Miner can effectively reduce the time and memory consumption in the mining process, thus providing data support for differentiated and precise maintenance strategies.

A novel diagnostic framework based on vibration image encoding and multi-scale neural network

Intelligent fault diagnosis employing CNN-based techniques has demonstrated promising results in rotating machinery maintenance and management. However, most approaches that rely on time-series vibration signals fail to incorporate the physical knowledge of faults into feature learning, and are subject to obtaining the correlation between input and output characteristics. Furthermore, it is challenging for the neural network to improve feature extraction due to the rigid convolution kernel and the limited receptive field. To address these issues, an intelligent fault diagnosis framework is developed based on vibration image encoding and multi-scale neural networks. Firstly, a novel image encoding rule is designed to convert time-series signals into vibration images, which incorporate the impact characteristics of faults into the model inputs to enhance fault information. Secondly, a multi-scale feature fusion network with dilatated and irregular sampling convolution block is constructed for feature extraction and classification. In addition, a new regularization term is designed to avoid training overfitting. Finally, Grad-CAM is introduced to visualize the region of interest of the model. The proposed method is verified on the datasets with different fault types from our laboratory and DIRG bearing dataset with different fault severities from Politecnico di Torino. The experimental results demonstrate that our diagnostic framework exhibits good generalization and noise robustness, and can achieve high-precision fault diagnosis.

Signal processing collaborated with deep learning: An interpretable FIRNet for industrial intelligent diagnosis

Due to the neglect of prior characteristics and the lack of explicit constraints on fault knowledge, conventional intelligent diagnosis methods suffer from great hardships in exacting fault-sensitive information and making explainable decisions, resulting in poor interpretability and inferior robustness. Motivated by the excellent multiscale analysis performance of signal processing and the powerful feature mining ability of deep learning, this study proposes an interpretable FIRNet for equipment intelligent diagnosis under strong noise environments. It consists of a well-crafted FIRLayer and a deep learning backbone. Inspired by the modulation principle that fault-sensitive components are normally modulated into multiscale mode characteristics, multiple ex-ante interpretable filters with two learnable parameters, including center frequency and bandwidth, are analytically designed to process the sequence signal sets, represented as signal-processing-based FIRLayer where the extracted multiscale feature maps are taken as an interpretable status information expression. Subsequently, multiscale convolutional kernels are established to extract the high-level feature maps and further make the final diagnostic decisions, represented as the deep learning backbone. The simulated and experimental results show that the proposed FIRNets have higher identification precision compared to the other nine deep learning models. Specially, three aspects, including model interpretability, noise robustness capacity and edge intelligent diagnosis, are further analyzed to illustrate the interpretable advantages of FIRNets. Hereinto, feature visualizations of the FIRLayer and mode decision contributions of the deep learning backbone are in-depth analyzed to verify the interpretable feature representation and decision-making principle of FIRNets. The results indicate that FIRNets have superior ex-post interpretability compared to other methods. Focusing on industrial practices, an efficient edge diagnosis system based on a pruned FIRNet is established, and an online diagnosis accuracy of more than 99% has been achieved. It can be foreseen that the proposed FIRNets show great potential and competitiveness to promote the edge computing application of equipment intelligent diagnosis.

Adaptive fractional-order integral fast terminal sliding mode and fault-tolerant control of dual-arm robots

AbstractClosed-loop kinematics of a dual-arm robot (DAR) often induces motion conflict. Control formulation is increasingly difficult in face of actuator failures. This article presents a new approach for fault-tolerant control of DARs based on advanced sliding mode control. A comprehensive fractional-order model is proposed taking nonlinear viscous and viscoelastic friction at the joints into account. Using integral fast terminal sliding mode control and fractional calculus, we develop two robust controllers for robots subject to motor faults, parametric uncertainties, and disturbances. Their merits rest with their strong robustness, speedy finite-time convergence, shortened reaching phase, and flexible selection of derivative orders. To avoid the need for full knowledge of faults, robot parameters, and disturbances, two versions of the proposed approach, namely adaptive integral fractional-order fast terminal sliding mode control, are developed. Here, an adaptation mechanism is equipped for estimating a common representative of individual uncertainties. Simulation and experiment are provided along with an extensive comparison with existing approaches. The results demonstrate the superiority of the proposed control technique. The robot performs well the tasks with better responses (e.g., with settling time reduced by at least 16%).

Gradient-based domain-augmented meta-learning single-domain generalization for fault diagnosis under variable operating conditions

Equipment operating conditions, referred to as domains, can induce domain drift in monitoring data, affecting data-driven fault diagnosis. Researchers have explored multi-domain generalization methods to tackle this issue. However, in actual industrial scenarios, the availability of fault data may be limited to a specific condition due to the cost or feasibility constraints associated with collecting extensive monitoring data. This limitation hampers the generalization ability of these methods, posing a major challenge for robust fault diagnosis under variable operating conditions. To address this challenge, we proposed a gradient-based domain-augmented meta-learning (GDM) single-domain generalization method. We analyze the restrictions of generating fake domains and construct a domain-augmented loss by evaluating diagnostic tasks minimization, semantic consistency, and distribution diversity for fake samples. Using a gradient-based technique, fake domains are generated iteratively, providing diverse fault knowledge for improved generalization. Instead of using time-consuming ensemble methods, we develop a novel meta-learning method to train a highly efficient and generalizable model, relaxing the requirement for auxiliary datasets in existing meta-learning methods. Two case studies consistently demonstrate the effectiveness and superiority of the proposed GDM method. Our findings suggest that this study offers a promising and competitive solution for single-domain generalization in fault diagnosis within real industrial scenarios.

A zero-shot fault attribute transfer learning method for compound fault diagnosis of power circuit breakers

Diagnosis of compound mechanical faults for power circuit breakers (CBs) is a challenging task. In traditional fault diagnosis methods, however, all fault types need to be collected in advance for the training of diagnosis model. Such processes have poor generalization capabilities for industrial scenarios with no or few data when faced with new faults. In this study, we propose a novel zero-shot learning method named DSR-AL to address this problem. An unsupervised neural network, namely, depthwise separable residual convolutional neural network, is designed to directly learn features from 3D time-frequency images of CB vibration signals. Then we build fault attribute learners (ALs), for transferring fault knowledge to the target faults. Finally, the ALs are used to predict the attribute vector of the target faults, thus realizing the recognition of previously unseen faults. The orthogonal experiments are designed and conducted on real industrial switchgear to validate the effectiveness of the proposed diagnosis framework. Results show that it is feasible to diagnose target faults without using their samples for training, which greatly saves the costs of collecting fault samples. This will help to accurately identify the various faults that may occur during CB’s life cycle, and facilitate the application of intelligent fault diagnosis system.

Fault Knowledge Graph Construction and Platform Development for Aircraft PHM.

To tackle the problems of over-reliance on traditional experience, poor troubleshooting robustness, and slow response by maintenance personnel to changes in faults in the current aircraft health management field, this paper proposes the use of a knowledge graph. The knowledge graph represents troubleshooting in a new way. The aim of the knowledge graph is to improve the correlation between fault data by representing experience. The data source for this study consists of the flight control system manual and typical fault cases of a specific aircraft type. A knowledge graph construction approach is proposed to construct a fault knowledge graph for aircraft health management. Firstly, the data are classified using the ERNIE model-based method. Then, a joint entity relationship extraction model based on ERNIE-BiLSTM-CRF-TreeBiLSTM is introduced to improve entity relationship extraction accuracy and reduce the semantic complexity of the text from a linguistic perspective. Additionally, a knowledge graph platform for aircraft health management is developed. The platform includes modules for text classification, knowledge extraction, knowledge auditing, a Q&A system, and graph visualization. These modules improve the management of aircraft health data and provide a foundation for rapid knowledge graph construction and knowledge graph-based fault diagnosis.

Dynamics simulation-driven fault diagnosis of rolling bearings using security transfer support matrix machine

Transfer fault diagnosis holds paramount significances in safeguarding the reliability and safety of rolling bearings. However, the current studies require massive experiment or field data in the source domain. Besides, sparse data (only single or several samples for each fault) and strong-noise data in the target domain are prone to cause the negative transfer problem. Importantly, most transfer fault diagnosis methods conduct fault classification in vector space, so the learning ability is weak for matrix-form fault features such as 2D time-frequency images. To address these issues, this paper proposes a novel dynamics simulation-driven fault diagnosis framework with security transfer support matrix machine (STSMM). In this framework, a high-fidelity bearing dynamic model is designed to generate sufficient source-domain data, which greatly reduces the acquisition cost of real data. A new matrix-form transfer learning model, i.e., STSMM, is proposed to effectively utilize the structural information contained in matrix data and achieve the simulation-to-real transfer of fault knowledge. Besides, a security transfer strategy is embedded in STSMM to improve the cross-domain diagnosis performance and theoretically avoid the negative transfer problem caused by sparse data and strong-noise data in the target domain. Experiment results demonstrate the effectiveness and superiority of the proposed framework.

Dynamic Optimization Method of Knowledge Graph Entity Relations for Smart Maintenance of Cantilever Roadheaders

The fault maintenance scenario in coal-mine equipment intelligence is composed of videos, images, signals, and repair process records. Text data are not the primary data that reflect the fault phenomenon, but rather the secondary processing based on operation experience. Focusing on the difficulty of extracting fault knowledge from the limited textual maintenance process records, a forward static full-connected topology network modeling method based on domain knowledge from four dimensions of physical structure, internal association, condition monitoring, and fault maintenance, is proposed to increase the efficiency of constructing a fault-maintenance knowledge graph. Accurately identifying the intrinsic correlation between the equipment anomalies and the faults’ causes through only domain knowledge and loosely coupled data is difficult. Based on the static full-connected knowledge graph of the cantilever roadheader, the information entropy and density-based DBSCAN clustering algorithm is used to process and analyze many condition-monitoring historical datasets to optimize the entity relationships between the fault phenomena and causes. The improved DBSCAN algorithm consists of three stages: firstly, extracting entity data related to fault information from the static fully connected graph; secondly, calculating the information entropy based on the real dataset describing the fault information and the historical operating condition, respectively; and thirdly, comparing the entropy values of the entities and analyzing the intrinsic relationship between the fault phenomenon, the operating condition data, and the fault causes. Based on the static full-connected topology storage in the Neo4j database, the information entropy and density-based DBSCAN algorithm is computed by using Python to identify the relationship weights and dynamically display optimized knowledge graph topology. Finally, an example of EBZ200-type cantilever roadheader for smart maintenance is studied to analyze and evaluate the forward and four-mainlines knowledge graph modeling method and the dynamic entity relations optimization method for static full-connected knowledge graph.

Multi-adversarial deep transfer network for multi-source open-set fault diagnosis of rotating machinery with category shift

Multi-source open-set fault diagnosis (MS-OSFD) with category shift is a critical challenge in industrial scenarios, as it involves potential unknown faults in the target domain and diverse label spaces across multiple source domains. The existing OSFD methods ignore the challenges posed by multiple source knowledge and category shift, resulting in incomplete fault knowledge transfer and subpar fault diagnosis performance. Aiming at the abovementioned problems, a multi-adversarial deep transfer network (MADTN) for MS-OSFD with category shift of rotating machinery is proposed. First, a multi-source fault classes matching (MSFCM) module is designed to align the data distribution of multiple source domains and alleviate the influence of category shift. Second, adversarial learning is adopted to reduce the domain gap between each pair of the source domain and target domain, and multiple fault identifiers integrated decision (MFIID) strategy is developed to obtain comprehensive predictions of target domain fault samples. Third, a global alignment (GA) is proposed based on the predictions of MFIID to further improve the domain-invariant feature extraction between domains. Finally, extensive experimental results demonstrate the superiority of the proposed MADTN method.