Fault Diagnosis Cases Research Articles

Intelligent fault diagnosis for unbalanced battery data using adversarial domain expansion and enhanced stochastic configuration networks

An accurate and efficient fault diagnosis method for battery systems is crucial to ensuring the safety of battery packs. Addressing the issue of insufficient actual fault data in battery operations, this paper proposes an intelligent fault diagnosis method based on feature-enhanced stochastic configuration networks and adversarial domain expansion of imbalanced battery fault data (AFDEM-FESCN). Firstly, we designed an adversarial fault domain data expansion method (AFDEM). By learning the distribution of fault data through adversarial training, the distribution of sample domains is balanced, thereby reducing model bias. Subsequently, we adjusted the distribution of SCN iterative parameters and added a linear feature layer. This enhances the feature extraction capability of the network through distribution overlay, enabling fault diagnosis. Finally, the effectiveness and feasibility of the proposed method were validated through a practical battery system fault diagnosis case, achieving a diagnostic accuracy of 92.1%. Experimental results demonstrate that the AFDEM-FESCN method exhibits good accuracy in battery system fault diagnosis, providing an effective solution to the challenge of imbalanced data in intelligent fault diagnosis.

Enhancing interpretability in neural networks for nuclear power plant fault diagnosis: A comprehensive analysis and improvement approach

Deep neural networks, as applied in the field of nuclear power fault diagnosis, have garnered significant attention alongside advancements in artificial intelligence technology. However, the “black box” nature of deep learning models has raised concerns regarding their deployment in scenarios demanding high safety standards, such as nuclear power plants. In this paper, we innovatively propose the utilization of an explainable artificial intelligence method, grounded in game theory, to conduct a detailed analysis of the diagnostic behavior of neural network models applied to nuclear power plants. By leveraging SHAP, the decision-making processes of these opaque models are demystified, offering insights into how and why they arrive at their predictions. In this study, two of the most widely utilized neural network frameworks are applied across six representative fault diagnosis cases for a comprehensive analysis. Based on its analysis results, a novel SHAP-Enhanced Feature Selection for Efficient Neural Network Fault Diagnosis strategy is proposed, significantly reducing model complexity without sacrificing diagnostic performance. This research breaks through the limitations of data-driven models used in the field of nuclear power fault diagnosis, conducts an interpretable analysis of their behavior, and proposes an improved strategy based on the analysis results, contributing to the practical application of data-driven models in nuclear power.

Reinforced fuzzy domain adaptation: Revolutionizing data-unaccessible rotating machinery fault diagnosis across multiple domains

Existing domain adaptation methods typically require access to raw data from both the source and target domains, increasing the risk of data leakage. In light of this, a fuzzy domain adaptation-based rotation machinery fault diagnosis method that does not require access to data is proposed in this paper. Specifically, during the source domain training phase, all source models are simultaneously trained by contributing model parameters. The developed joint training strategy significantly enhances the multi-domain performance of each private source model. Simultaneously, a model incorporating a fusion fuzzy strategy is constructed to acquire fuzzy rules for the source domain samples and anchor points corresponding to each fault mode. During the target domain adaptation phase, firstly, under the setting of frozen fuzzy rules, the source feature extractor is retrained using target data. Subsequently, an anchor point alignment strategy is introduced to transfer diagnostic knowledge between domains. Finally, a filtering strategy is devised to obtain high-confidence pseudo-labels for the target domain, coupled with a self-supervised learning strategy for the final optimization of the target model. Experimental results demonstrate significant achievements of the proposed method across twelve multi-source cross-domain fault diagnosis cases, with an average diagnostic accuracy of 98.88%. Encouragingly, the proposed method maintains outstanding performance even without accessing source data, outperforming several other methods.

The Potentiality of Integrating Model-Based Residuals and Machine-Learning Classifiers: An Induction Motor Fault Diagnosis Case

The Potentiality of Integrating Model-Based Residuals and Machine-Learning Classifiers: An Induction Motor Fault Diagnosis Case

Causal Disentanglement Domain Generalization for time-series signal fault diagnosis

Domain generalization-based fault diagnosis (DGFD) presents significant prospects for recognizing faults without the accessibility of the target domain. Previous DGFD methods have achieved significant progress; however, there are some limitations. First, most DGFG methods statistically model the dependence between time-series data and labels, and they are superficial descriptions to the actual data-generating process. Second, most of the existing DGFD methods are only verified on vibrational time-series datasets, which is insufficient to show the potential of domain generalization in the fault diagnosis area. In response to the above issues, this paper first proposes a DGFD method named Causal Disentanglement Domain Generalization (CDDG), which can reestablish the data-generating process by disentangling time-series data into the causal factors (fault-related representation) and no-casual factors (domain-related representation) with a structural causal model. Specifically, in CDDG, causal aggregation loss is designed to separate the unobservable causal and non-causal factors. Meanwhile, the reconstruction loss is proposed to ensure the information completeness of the disentangled factors. We also introduce a redundancy reduction loss to learn efficient features. The proposed CDDG is verified on five cross-machine vibrational fault diagnosis cases and three cross-environment acoustical anomaly detection cases by comparing it with eight state-of-the-art (SOTA) DGFD methods. We publicize the open-source time-series DGFD Benchmark containing CDDG and the eight SOTA methods. The code repository will be available at https://github.com/ShaneSpace/DGFDBenchmark.

Smart filter aided domain adversarial neural network for fault diagnosis in noisy industrial scenarios

The application of unsupervised domain adaptation (UDA)-based fault diagnosis methods has shown significant efficacy in industrial settings, facilitating the transfer of operational experience and fault signatures between different operating conditions, different units of a fleet or between simulated and real data. However, in real industrial scenarios, unknown levels and types of noise can amplify the difficulty of domain alignment, thus severely affecting the diagnostic performance of deep learning models. To address this issue, we propose an UDA method called Smart Filter-Aided Domain Adversarial Neural Network (SFDANN) for fault diagnosis in noisy industrial scenarios. The proposed methodology comprises two steps. In the first step, we develop a smart filter that dynamically enforces similarity between the source and target domain data in the time-frequency domain. This is achieved by combining a learnable wavelet packet transform network (LWPT) and a traditional wavelet packet transform module. In the second step, we input the data reconstructed by the smart filter into a domain adversarial neural network (DANN). To learn domain-invariant and discriminative features, the learnable modules of SFDANN are trained in a unified manner with three objectives: time-frequency feature proximity, domain alignment, and fault classification. We validate the effectiveness of the proposed SFDANN method based on two fault diagnosis cases: one involving fault diagnosis of bearings in noisy environments and another involving fault diagnosis of slab tracks in a train-track-bridge coupling vibration system, where the transfer task involves transferring from numerical simulations to field measurements. Results show that compared to other representative state of the art UDA methods, SFDANN exhibits superior performance and remarkable stability.

Explainable Deep Ensemble Model for Bearing Fault Diagnosis Under Variable Conditions

Deep learning-based intelligent diagnostic methods have been widely used in aerospace, rail transportation, automotive, rail vehicles, and other fields. However, deep neural networks are composed of multi-layer non-linear transformation to learn the feature maps of input, leading to the black box characteristic. It is usually criticized to be non-transparent and less interpretable in predictions and maintenance decision-making. Meanwhile, due to the variability of machine working conditions in the manufacturing process, there is still a gap between traditional deep learning diagnosis methods and the practical application. To address such problems, an explainable deep ensemble model is proposed for fault diagnosis of bearings which both leverages feature learning ability and improves model interpretability. Firstly, a residual network integrated with bi-directional long short-term memory (BiLSTM) is designed as the backbone learning framework to perform associative feature extraction from the original 1D time-series signals. Then, a multi-branch model with a multi-scale ensemble learning strategy is constructed for feature fusion and final fault diagnosis. After the model is trained, an enhanced gradient-weighted class activation mapping technique, named 1D Grad-CAM ++ is embedded to visualize the region of interest in the convolutional layer, where the prior knowledge gained from the bearing fault mechanism model is effectively combined to provide a better understanding of the learned features and decision-making of model. Finally, three fault diagnosis cases of rolling bearings under different operation conditions are used for algorithm verification. The results demonstrate the proposed approach shows superior explainable capability and diagnosis performance in comparison with other methods.

Unbalanced Bearing Fault Diagnosis Under Various Speeds Based on Spectrum Alignment and Deep Transfer Convolution Neural Network

Bearing fault diagnosis plays a pivotal role in the safe and reliable operation of modern mechanical systems. However, the existing fault diagnosis methods rarely deal with the problem of category imbalance and various speeds concurrently, which cannot work effectively in practical scenarios. Considering the underlying similarities of data in frequency domain, data mining under various speeds can help to reduce the deviation of domain distribution. Therefore, a novel fault diagnosis method based on spectrum alignment (SA) and deep transfer convolution neural network (DTCNN) is proposed, where the SA and data augmentation module (DAM) are designed to extract spectrum alignment features from the unbalanced bearing data. The DTCNN model based on joint distribution adaptation (JDA) is built to facilitate learning reliable domain-invariant features. Different from the existing studies, a more general transfer task with time-varying speed is considered, even with complex faults. For 14 transfer tasks in two unbalanced fault diagnosis cases under variable speed, the average accuracy, F1-score, area under curve (AUC) of the proposed method can reach more than 97.76% 97.57% 98.75%. The results show that this method has superior diagnostic effect and better generalization ability than various state-of-the-art methods.

Fault Diagnosis for Complex Equipment Based on Belief Rule Base with Adaptive Nonlinear Membership Function

Fault diagnosis of complex equipment has become a hot field in recent years. Due to excellent uncertainty processing capability and small sample problem modeling capability, belief rule base (BRB) has been widely used in the fault diagnosis. However, previous BRB models almost did not consider the diverse distributions of observation data which may reduce diagnostic accuracy. In this paper, a new fault diagnosis model based on BRB is proposed. Considering that the previous triangular membership function cannot address the diverse distribution of observation data, a new nonlinear membership function is proposed to transform the input information. Then, since the model parameters initially determined by experts are inaccurate, a new parameter optimization model with the parameters of the nonlinear membership function is proposed and driven by the gradient descent method to prevent the expert knowledge from being destroyed. A fault diagnosis case of laser gyro is used to verify the validity of the proposed model. In the case study, the diagnosis accuracy of the new BRB-based fault diagnosis model reached 95.56%, which shows better fault diagnosis performance than other methods.

A Lifelong Learning Method Based on Generative Feature Replay for Bearing Diagnosis With Incremental Fault Types

Deep learning (DL)-based fault diagnosis models need to collect sufficient fault information for each fault type to ensure high-precision diagnosis. Some unexpected and new fault types will inevitably appear in actual conditions, which are called incremental fault types or class increment. Traditional DL models require the costly collection of all known data for retraining; while the use of new fault data may lead to catastrophic forgetting of old tasks. To solve the problem of bearing diagnosis with incremental fault types, a lifelong learning method based on generative feature replay (LLMGFR) is proposed in this study. A feature distillation method is put forward in this method to avoid forgetting in the feature extractor. The generator is trained to produce old task features. The generated features are mixed with real features of the current task to solve the imbalance problem and catastrophic forgetting of the classifier effectively. According to incremental fault diagnosis cases, LLMGFR can learn constantly and adaptively in dynamic environments with incremental fault types.

Robustan algoritam za učenje pravila za klasifikaciju - studija slučaja dijagnoze greške

Machine learning algorithms are used for building classifier models. The rule-based decision tree classifiers are popular ones. However, the performance of the decision tree classifier varies with hyperparameter tuning. The optimum hyperparameter values are obtained using either optimization algorithms or trial and error methods. The present study utilizes the MODLEM algorithm to overcome the drawbacks accounted for by decision tree algorithms. Eliminating hyperparameter tuning and producing results closer to standard decision tree algorithms makes MODLEM a robust classification algorithm. The robustness of the MODLEM algorithm is illustrated with the fault diagnosis case study. The case study is faults diagnosis of an automobile suspension system using vibration signals acquired at various fault conditions.

Group-Sparse Feature Extraction via Ensemble Generalized Minimax-Concave Penalty for Wind-Turbine-Fault Diagnosis

Extracting weak fault features from noisy measured signals is critical for the diagnosis of wind turbine faults. In this paper, a novel group-sparse feature extraction method via an ensemble generalized minimax-concave (GMC) penalty is proposed for machinery health monitoring. Specifically, the proposed method tackles the problem of formulating large useful magnitude values as isolated features in the original GMC-based sparse feature extraction method. To accurately estimate group-sparse fault features, the proposed method formulates an effective unconstrained optimization problem wherein the group-sparse structure is incorporated into non-convex regularization. Moreover, the convex condition is proved to maintain the convexity of the whole formulated cost function. In addition, the setting criteria of the regularization parameter are investigated. A simulated signal is presented to verify the performance of the proposed method for group-sparse feature extraction. Finally, the effectiveness of the proposed group-sparse feature extraction method is further validated by experimental fault diagnosis cases.

Analysis and Research on Vibration Characteristics of Nuclear Centrifugal Pumps Rotor Rubbing Fault

To investigate the vibration characteristics of nuclear centrifugal pumps rotor rubbing fault, the mechanical model of rotor rub is established, and the causes, types, and mechanisms of rotor rub fault are introduced. Taking the HAZ50-250 Type Horizontal single-stage centrifugal pump prototype as the object, through a reasonable test scheme, the vibration test and analysis of the rotor under different rub conditions were carried out using the Enpac2500 Dual channel spectrum analyzer, and the fault data of different rub impact degrees were collected. Combining the characteristics of vibration parameters such as time-domain, frequency-domain, and axis trajectory, it is found that the time-domain signal of rub fault shows an obvious “Top Clipping phenomenon.” The frequency-domain signal is mainly concentrated in one-time rotation frequency, two times rotation frequency, and a large number of nonlinear high harmonics. The axis trajectory will produce a concave image in the rub direction. In the actual rotor rub vibration fault diagnosis, it is necessary to comprehensively judge the vibration caused in combination with the operating state of the centrifugal pump, the state of the time domain, the frequency domain, and the axis trajectory. Based on the analyzed spectrum characteristics of the rub impact vibration, it is applied to the vibration fault diagnosis case of the centrifugal pump in the power plant, which timely prevents the pump damage and other faults caused by the deterioration of the rub impact fault, and plays a key role in the safe and stable operation of the centrifugal pump. It also provides an important reference for pump design and vibration fault diagnosis of rotating equipment in nuclear power plants.

Refined composite multiscale fuzzy entropy based fault diagnosis of diesel engine

Due to complicated transfer paths and strong background noise interference, the fault pattern information deeply hides in common features of the vibration signal at the engine surface. In this study, the refined composite multiscale fuzzy entropy (RCMFE) used to measure the irregularity and self-similarity of time series is proposed to quantify the feature of various fault patterns. Followed by RCMFE, the features dug out are recognized by a parameter-adaptive support vector machine based on the firefly algorithm (FASVM). After putting forward the diagnosing schematics, the RCMFE-FASVM is applied to a fault diagnosis case of a diesel engine on a test rig. A comparative analysis of the four methods to extract features and the four methods to recognize fault patterns are conducted. Results indicate the proposed method has superior recognition performance and can effectively identify the working states of the diesel engine, contrasting the existing methods. Under the small samples and features task of identifying the working states of a diesel engine, the recognition rate of the proposed method with more stability can reach 98.2%, which is larger than other methods. Given the superior performance of the proposed method, the number of input features and training samples should vary from 8 to 20 and from 35 to 50.

High-temperature augmented neighborhood metric learning for cross-domain fault diagnosis with imbalanced data

The class imbalance in the machine monitoring data and the data distribution discrepancy due to the different working conditions make it difficult to build a strong mechanical fault diagnosis model in practice. In this paper, we proposed a high-temperature augmented neighborhood metric-learning network (HNMN) for cross-domain fault diagnosis with imbalanced data. In the proposed network, fault types are classified by comparing the distance from the sample to the prototype. And neighborhood component analysis-based domain adaptation scheme was presented to reduce the distance of similar samples in source and target domains thus eliminating the domain shift and solving the variable working condition problem. A classification boundary optimization strategy based on high-temperature mechanism was applied to avoid the overlap of feature distributions of minor classes and major classes, and finally solve the class imbalance problem. The performance and superiority of HNMN are evaluated and analyzed using three fault diagnosis cases. The experimental results demonstrate that the proposed network can achieve higher fault diagnosis accuracy compared to the state-of-the-art methods in the class imbalance and variable working condition diagnosis scenarios.

Continual learning fault diagnosis: A dual-branch adaptive aggregation residual network for fault diagnosis with machine increments

As a data-driven approach, Deep Learning (DL)-based fault diagnosis methods need to collect the relatively comprehensive data on machine fault types to achieve satisfactory performance. A mechanical system may include multiple submachines in the real-world. During condition monitoring of a mechanical system, fault data are distributed in a continuous flow of constantly generated information and new faults will inevitably occur in unconsidered submachines, which are also called machine increments. Therefore, adequately collecting fault data in advance is difficult. Limited by the characteristics of DL, training existing models directly with new fault data of new submachines leads to catastrophic forgetting of old tasks, while the cost of collecting all known data to retrain the models is excessively high. DL-based fault diagnosis methods cannot learn continually and adaptively in dynamic environments. A new Continual Learning Fault Diagnosis method (CLFD) is proposed in this paper to solve a series of fault diagnosis tasks with machine increments. The stability–plasticity dilemma is an intrinsic issue in continual learning. The core of CLFD is the proposed Dual-branch Adaptive Aggregation Residual Network (DAARN). Two types of residual blocks are created in each block layer of DAARN: steady and dynamic blocks. The stability–plasticity dilemma is solved by assigning them with adaptive aggregation weights to balance stability and plasticity, and a bi-level optimization program is used to optimize adaptive aggregation weights and model parameters. In addition, a feature-level knowledge distillation loss function is proposed to further overcome catastrophic forgetting. CLFD is then applied to the fault diagnosis case with machine increments. Results demonstrate that CLFD outperforms other continual learning methods and has satisfactory robustness.

Multisource cross-domain fault diagnosis of rolling bearing based on subdomain adaptation network

The excellent performance of current intelligent fault diagnosis methods based on deep learning is attributed to the availability of large amounts of labeled data. However, in practical bearing fault diagnosis, the high cost of large sample data and changes in operating conditions lead to the scarcity of available training data, which limits the engineering application of intelligent bearing fault diagnosis. To solve this problem, this paper proposes a cross-domain fault diagnosis method based on multisource subdomain adaptation networks (MSDAN). First, the data from multiple source domains are simultaneously input to a shared feature extractor composed of a one-dimensional residual network. Then, the private feature extractor is used to learn features from different source domains and reduce the domain shifts of each source and target domain using the local maximum mean discrepancy. Finally, the different classifier outputs of the target domain samples are aligned. The highlight of MSDAN is to obtain diagnostic knowledge from multiple source domains and further divide the subdomains using the categories as criteria, which not only aligns the global distribution of the source and target domain but also performs a more refined subdomain alignment. The method effectively alleviates the negative transfer phenomenon caused by insufficient domain alignment in multisource transfer diagnosis. The effectiveness and superiority of the proposed MSDAN method are verified by constructing seven multisource transfer tasks with two bearing fault diagnosis cases, including cross-operating-condition and cross-machine.

A fault diagnosis method for the HIPPS of FPSO unit based on dynamic Bayesian network

The high integrity pressure protection system (HIPPS) on the Floating Production Storage and Offloading (FPSO) unit is essential for handling various emergencies. However, if the location of the fault cannot be accurately identified, proper measures may not be taken to isolate the hazard. This paper presents a fault diagnosis method for HIPPS on FPSO units based on Dynamic Bayesian network (DBN). The method considers the influence of sensor and system equipment degradation on the diagnosis results and avoids the problem of overdiagnosis in static diagnosis networks. Six fault diagnosis cases of the system are analyzed and discussed to verify the accuracy and effectiveness of the proposed method. By changing the failure rate of the faulty component, it is determined that the posterior probability of the faulty component increases with the increase of the failure rate at the same time.

Collaborative and adversarial deep transfer auto-encoder for intelligent fault diagnosis

Deep transfer learning provides an advanced analytical tool for intelligent fault diagnosis to learn shared fault knowledge in industrial scenarios whereby datasets are collected from different operating conditions. The majority of previous approaches consist in minimizing the domain discrepancy at the domain level, but they may cause erroneous mappings for local category features. Recently, to prioritize the feature discriminability, several adversarial learning approaches with bi-classifier have been developed by using the classification disagreement. However, they suffer from a restricted representation of the classification disagreement, and may fail to align category-level features when domain gaps are large. To address these weaknesses, we proposed a Collaborative and Adversarial Deep Transfer model based on a convolutional Auto-encoder (CADTA) for intelligent fault diagnosis. Specifically, by leveraging a couple of multi-task classifiers, a joint subspace feature discrimination method involving duplex adversarial learning is proposed to promote the category-level feature discriminability. Then, a collaborative transfer scheme is built to integrate the domain similarity learning and the joint subspace feature discrimination hierarchically into a convolutional Auto-encoder, in which the domain similarity learning is achieved by adapting the intermediate feature representations in terms of high-order moments. For the sake of efficient model training, a stage-wise adversarial training process is presented correspondingly. Extensive experiments of diverse transfer tasks based on two rolling bearing datasets and three transfer fault diagnosis cases demonstrate that CADTA outperforms the existing state-of-the-art deep transfer learning methods.

Bearing Fault Diagnosis Under Small Data Set Condition: A Bayesian Network Method With Transfer Learning for Parameter Estimation

Bearings are broadly applied in various types of industrial systems. Fault diagnosis, as a promising way for reliability of modern industrial internet of thing applications, has attracted increasing attention from both academia and industry fields. Being ideal modeling and inference tool in uncertainty situations, Bayesian network (BN) is becoming increasingly popular in many systems. However, in practical uncertain and complicated engineering surroundings, it’s difficult or expensive to collect massive labeled fault data for the sake of fault diagnosis model learning. To address the issue of BN parameter learning under small data set conditions, this paper proposes a Varying Coefficient Transfer Learning (VCTL) algorithm based on aggregation and transfer learning, that considers both knowledge from the resource domain and the resource relevance contributions. The balancing weight function is designed to determine whether the learning task in the resource domain is activated. Relevance weight factors are proposed to measure the relevance of resource and target parameters quantitatively, by combing parameter information from resource domains with those obtained from the target domain, using maximum a posterior (MAP) or maximum likelihood estimation (MLE). Finally, the target parameters are aggregated with both the target initial parameters and the parameter knowledge from the resource domain. Based on VCTL, a bearing fault diagnosis approach is proposed and verified. The experimental results show that, under the condition of the small data set, learning accuracy of VCTL algorithm with varying coefficient aggregation is better than MLE algorithm, MAP algorithm or state-of-the-art parameter transfer method, local linear pooling transfer learning (LoLP) algorithm. Under the condition of sufficient data set, learning accuracy of VCTL algorithm approaches the classical MLE or MAP, and the correctness of the proposed algorithm is verified. Moreover, we illustrate the successful application to real-world bearing fault diagnosis case with VCTL, where we had access to expert-provided resource knowledge and real fault diagnosis data.