Fault Diagnosis Methodology Research Articles

A feature selection and ensemble learning based methodology for transformer fault diagnosis

Dissolved gas analysis (DGA) data are generally used to diagnose a transformer fault. However, the measurement errors in DGA data are inevitable and will affect the accuracy and reliability of the diagnosis results. Nevertheless, so far, only a few efforts have been devoted to addressing this issue. To provide an accurate and stable transformer fault diagnosis system, a feature selection and ensemble learning based methodology is proposed. Firstly, an Overlapping Information Feature Selection (OIFS) method is proposed to select efficient features from the given feature set for classifiers. Secondly, an Intelligent Voting Ensemble Learning (IVEL) method is proposed to generate a diagnosis model based on the feature combination from the OIFS method. The reported test results show that the OIFS method outperforms existing methods in 9 out of 10 tested classifiers. Additionally, the IVEL outperforms three popular ensemble learning methods, including random forest (RF), gradient boosting decision tree (GBDT), and LightGBM, in terms of both accuracy and robustness performances. Finally, the proposed methodology (OIFS-IVEL) is applied to diagnose the transformer faults in the IEC TC 10 database, achieving a 100% accuracy in recognizing fault types and a 92.6 % accuracy in evaluating fault severity.

Selection of relevant features to detect and diagnose single and multiple simultaneous soft faults in air-source heat pumps

Soft faults in air-source heat pump systems can cause various issues, such as decreased performance, increased energy consumption, and long-term damage. This paper presents a comprehensive modeling approach of air-source heat pumps that considers individual and simultaneous soft faults validated against empirical experiments from the literature, a perspective limited in previous research. The methodology is based on the simulation of fault-free and faulty system performance utilizing heat pump modeling software. The results are a selection of variables and their trends to select the appropriate features to detect and diagnose single and double soft faults. The model also computes the impacts of these combined faults on COP, revealing significant reductions. Specifically, combinations of overcharge with condenser fouling and undercharge with compressor valve leakages induce COP reductions ranging from 17% to 38%, depending on operating conditions. The model can be used to develop faulty heat pump data, which can be adapted to other system typologies and conditions and used as input for data-driven Fault Detection and Diagnosis (FDD) methodologies.

Cross-validation enhanced digital twin driven fault diagnosis methodology for minor faults of subsea production control system

A subsea production system is essential for the subsea production of oil and gas. Real-time monitoring can ensure safe production. A subsea production control system is the core of the subsea production system and the top priority to be monitored. Minor faults refer to faults with weak characteristics and are difficult to be found. They are common and pose significant risks in subsea production system. Safe operation of the system can be guaranteed to the greatest extent with a timely detection and handling of minor faults. Digital Twin driven fault diagnosis is an effective method to monitor the subsea production control system. However, the combination of digital twin and fault diagnosis is not comprehensive, especially in data interaction. It leads to the fact that digital twin cannot improve fault diagnosis accuracy significantly, mainly when a minor fault occurs. A cross-validation enhanced digital twin-driven fault diagnosis methodology for minor faults of the subsea production control system is proposed to achieve high accuracy. Control, loss and fault parameters are presented and used for building a digital twin model. Bayesian Networks are used to construct a fault diagnosis model. Single and cumulative errors are used to measure the difference between digital twin models and physical systems. A verification feedback method is used to check the diagnosis results. Data of four days of a subsea production system in the South Sea of China is used to demonstrate the proposed methodology. The results show that the method can improve the diagnosis accuracy for minor faults effectively.

An efficient sensor and thermal coupling fault diagnosis methodology for building energy systems

An efficient sensor and thermal coupling fault diagnosis methodology for building energy systems

Generative adversarial network to alleviate information insufficiency in intelligent fault diagnosis by generating continuations of signals

This paper introduces Con-GAN, an innovative improvement of GAN-based data augmentation designed to address data insufficiency in fault diagnosis methodologies. Distinctly different from traditional GAN-based data augmentation, Con-GAN aims to generate realistic continuations of existing signals and subsequently integrates these continuations with the original signals. This ’real-fake-mixed’ strategy fully leverages the existing signals and results in high-quality new signals for data augmentation, making our approach both safer and more effective. With rigorous validation across multiple datasets, which include both artificially induced and test-caused real faults, Con-GAN’s consistent effectiveness has been substantiated. Compared to other GANs, Con-GAN presents distinct advantages, yielding new signals with better validity and variety. In summary, Con-GAN introduces a novel and feasible paradigm in GAN-based data augmentation, demonstrating practical value for industrial. In the future, we plan to evaluate Con-GAN’s performance under conditions with high noise levels, exploring its robustness and adaptability across a wider range of scenarios.

Large margin-learning methodology from time-frequency maps and its physically interpretable weights for simultaneous machine health monitoring and fault diagnosis

Artificial intelligence (AI) enables machine health monitoring including fault detection and diagnosis that are becoming prevalent and have been actively investigated during recent years. Nonetheless, poor interpretation and black-box characteristics hinder their performance in machine condition monitoring. Specifically, abstract features extracted from AI models are difficult to physically interpret and correlate with fault mechanisms. In this paper, to enrich interpretation of learned features, a large margin-learning methodology from time-frequency maps is newly proposed. The original idea is to construct a health indicator (HI) for degradation process tracking in a feature extraction module and it is defined as the sum of element-wise products between a time–frequency map and a weight matrix. Here, the weight matrix can indicate fundamental components that always exist in healthy and faulty signals and newly generated components that are frequency components caused by machine faults. Moreover, the significance of the two-dimensional weight matrix can clearly show how weights are varied with time and indicate a cyclo-stationarity behavior of repetitive transients existing in fault vibration signals, which results in a new perspective for revisiting time–frequency maps for intelligent large margin-learning. Following, thanks to the physically interpretable feature matrix, a single-hidden layer network in a decision-making module is seamlessly designed for immediate machine fault detection, in which weights between an input layer and a hidden layer are tactfully set to the physically interpretable feature. At last, since the proposed methodology is based on convex optimization, it only needs a small amount of samples for training and realizing the physically interpretable weight matrix to understand a significant difference between different time–frequency maps for machine health monitoring and diagnosis. Experiments verify the effectiveness and superiority of the proposed large margin-learning methodology for simultaneous machine health monitoring and fault diagnosis.

Incipient fault diagnosis and trend prediction in nonlinear closed-loop systems with Gaussian and non-Gaussian noise

This paper proposes a methodology for incipient fault diagnosis and the corresponding trend prediction in nonlinear closed-loop systems considering stochastic Gaussian and non-Gaussian uncertainties. The proposed approach is based on the use of the particle filtering technique for estimating the system states and outputs. From these estimations, the residual signals would be generated through a mathematical filtering and augmentation technique, allowing the incipient fault estimation that is evaluated using the designed fixed and adaptive thresholds that consider system uncertainties. In this way, the fault detection performance is improved but also the false detection and false alarm problems are comprehensively addressed. Moreover, the augmented Gauss–Newton identification method is used for the incipient fault trend prediction. Finally, to evaluate the effectiveness of the proposed approach, the incipient fault diagnosis in the heat transfer unit built in the nonlinear closed-loop continuous stirred-tank reactor (CSTR) system is used. Besides, the confusion matrix is employed to assess the results from a quantitative point of view.

Entropy-maximization oriented interpretable health indicators for locating informative fault frequencies for machine health monitoring

Machine health monitoring is an important domain to provide timely anomaly detection and diagnostic supports for condition based maintenance. Health indicator (HI) construction is an intuitive and efficient way to conduct continuous machine health monitoring. In this study, a maximization-entropy optimization oriented interpretable HI is proposed to locate informative fault frequencies for machine health monitoring. At first, a probabilistic HI is defined based on the sum of weighted spectral amplitudes in normalized square envelope spectrum (NSES) and its newly defined constraint. Next, different from maximization of negative entropy or minimization of entropy for extracting repetitive transients caused by localized faults, this paper provides a new perspective for integrating entropy maximization with maximum sparsity weights to propose a novel convex optimization degradation modeling methodology for construction of HIs. Besides the new perspective of entropy maximization, the novelty of this study lies in that optimized model weights can automatically locate fault characteristic frequencies in the NSES without any prior knowledge so that the probabilistic HI is sensitive to incipient abnormality. Based on the analysis of several run-to-failure bearing datasets, it is demonstrated the effectiveness and superiority of the proposed methodology for machine health monitoring and incipient fault diagnosis.

A novel hybrid methodology for fault diagnosis of wind energy conversion systems

This paper proposes effective Random Forest (RF)-based fault detection and diagnosis for wind energy conversion (WEC) systems. The proposed technique involved two major steps: feature selection and fault classification. Feature selection pre-processing is an important step to increase the accuracy of the classification algorithm and decrease the dimensionality of a dataset. Therefore, a hybrid feature selection based diagnosis technique, that can preserve the advantages of wrapper and filter algorithms as well as RF model, is proposed. In the first phase, the neighborhood component analysis (NCA) filter algorithm is used to reduce and select only the pertinent features from the original raw data. This phase helps in improving data by removing redundant and unimportant features. In the second step, we applied a wrapper technique called equilibrium optimizer to get optimized features and better classification accuracy. The main idea behind using a hybrid feature selection step is to select a small subset from original data that can achieve maximum classification accuracy and reduce the computational complexity of the RF technique. Then, the sensitive and significant characteristics are transmitted to the RF model for classification purposes. The presented results prove that the proposed methods offer enhanced diagnosis accuracy when applied to WEC systems.

A cascade neural network methodology for fault detection and diagnosis in solar thermal plants

Detecting and isolating faults in collector fields of solar thermal power plants is a crucial and challenging task. The system variables in the collector area are highly coupled, which can lead to a high misclassification rate. For this reason, it becomes necessary to combine knowledge of systems engineering with machine learning techniques that unravel the complex dynamics that govern the systems using historical data. Furthermore, the performance of a solar thermal plant is highly dependent on solar irradiance which changes during the day and is subject to perturbations caused by clouds and other atmospheric conditions. Detecting the fault requires using techniques that cope with the disturbances in solar irradiance.In this work, real irradiance profiles with many types of clouds are used. First, a model-based fault detector is applied, obtaining an accuracy of over 89% for all test irradiances. Then, different machine learning techniques are compared: static neural networks with and without decoupling strategy, dynamic neural networks, dynamic neural networks in cascade, classification trees, random forests, radial basis function networks, and self-organizing maps. The combination of neural networks was the only method that obtained a total accuracy of over 73% and F1-scores over 50% for all the test irradiance profiles.

An enhanced temporal algorithm- coupled optimized adaptive sparse principal component analysis methodology for fault diagnosis of chemical processes

Principal component analysis (PCA) is a classic fault diagnosis method widely used in chemical process data modeling. However, the limitation of PCA to handle dynamic and time-variant data has been progressively exposed with the ever-increasing development of artificial intelligence technologies and applications of automation systems in the chemical industry. In this paper, an enhanced temporal algorithm-coupled optimized adaptive sparse PCA (ETA-OASPCA) methodology is proposed with the aim to improve the conventional PCA method by introducing temporal state computation and dynamic adaptive sparsity. The ETA is constructed to extract the state transitions of the time-variant process data by introducing the “Transition” and “Hold”, acquire the region of interest by employing a Maximum Fuzzy c-means (MFCM) method, and then use the region time prediction automaton to label the possible abnormal states. The introduction of dynamic adaptive sparsity in PCA, achieved by presenting an iterative interior point optimization method to update the dynamic sparse penalty term, can impose sparsity constraints on dynamic data calculations, thus enhancing the interpretability of the PCA model. The effectiveness of the proposed method was validated through its application in the Eastman Tennessee process, showing that an average fault detection rate, false alarm rate, and latency of 99.4%, 0.37%, and 4.11 mins can be achieved, respectively.

A methodology for transformer fault diagnosis based on the feature extraction from DGA data

A methodology to diagnose transformer faults based on dissolved gas analysis (DGA) is proposed. Since a fault is more sensitive to the ratios of gas contents, a general ratio feature extraction framework is proposed to generate a feature subset as the input of the diagnosis model. The feature subset is evaluated by the improved case-based reasoning (CBR) and optimized by using a proposed new algorithm called the k-optimal algorithm (k-OA). The comparison results between the k-OA and the genetic algorithm (GA) show that the k-OA is more efficient in solving such a combinatorial optimization problem. The obtained optimal feature subset is used to diagnose a public transformer fault dataset, and a 92.6% diagnosis accuracy is observed as compared to that of only 85.1% diagnosis accuracy by using the original features.

Erratum: “A variable-speed-condition bearing fault diagnosis methodology with recurrence plot coding and MobileNet-v3 model” [Rev. Sci. Instrum. 94, 034710 (2023)

First Page

Fault Diagnosis of Rotating Machinery With Limited Expert Interaction: A Multicriteria Active Learning Approach Based on Broad Learning System

Recently, research on the fault diagnosis of rotating machinery, especially for the compound or unknown cases, has drawn increasing attention. Some advanced learning-based approaches have achieved good fault diagnosis performance to some degree. However, in practical applications, how to utilize prior knowledge as fully as possible for fault diagnosis with constraints of limited expert interaction remains an open issue. In this brief, a fault diagnosis methodology of rotating machinery with limited expert interaction is proposed. With related feature extraction techniques, a novel multicriteria active learning (MCAL) query strategy is designed to select the relatively valuable samples for annotation. In addition, the broad learning system (BLS) is exploited to achieve fast incrementally updating or retrain procedures with high diagnostic accuracy in different diagnosis scenarios. Several experiments are conducted on a real-world rotating machinery fault diagnosis (RMFD) experimental platform. Compared with other existing advanced approaches, the diagnosis performance of the proposal shows high stability and flexibility. The annotation cost of experts is also significantly reduced, which makes the proposal more suitable for dealing with practical problems.

Fault Diagnosis Methodology of Redundant Closed-Loop Feedback Control Systems: Subsea Blowout Preventer System as a Case Study

In closed-loop feedback control systems, faults are propagated through the feedback link, which eventually leads to the abnormality of the entire system. Generally, it is very difficult to identify the faults of systems under the influence of the closed-loop feedback link. The existence of redundancy improves the reliability of the system. Meanwhile, it also poses new challenges to the fault diagnosis of multiple redundant systems. In this regard, a causality-based method is proposed for the fault diagnosis of closed-loop feedback control system with multiple modular redundancy. The dynamic Bayesian networks for fault diagnosis are established based on sensor data and system parameters. The networks consist of four layers, which are sensors, performances, monitors, and faults, respectively. Furthermore, the conditional probabilities of the fault nodes are calculated by Noisy-OR and Noisy-MAX models. The proposed method can dynamically evaluate system performance and integrate other monitoring information as evidence to assist faults diagnosis and location. A double modular redundant control system for a subsea blowout preventer is used as a case to demonstrate the proposed method, and the results show that the proposed method has high accuracy. The influence of sampling frequency, noise, and redundancy mode on diagnosis results is studied and discussed in the case study.

A variable-speed-condition bearing fault diagnosis methodology with recurrence plot coding and MobileNet-v3 model.

To improve the quality of the non-stationary vibration features and the performance of the variable-speed-condition fault diagnosis, this paper proposed a bearing fault diagnosis approach with Recurrence Plot (RP) coding and a MobileNet-v3 model. 3500 RP images with seven fault modes were obtained with angular domain resampling technology and RP coding and were input into the MobileNet-v3 model for bearing fault diagnosis. Additionally, we performed a bearing vibration experiment to verify the effectiveness of the proposed method. The results show that the RP image coding method with 99.99% test accuracy is superior to the other three image coding methods such as Gramian Angular Difference Fields, Gramian Angular Summation Fields, and Markov Transition Fields with 96.88%, 90.20%, and 72.51%, indicating that the RP image coding method is more suitable for characterizing variable-speed fault features. Compared with four diagnosis methods such as MobileNet-v3 (small), MobileNet-v3 (large), ResNet-18, and DenseNet121, and two state-of-the-art approaches such as Symmetrized Dot Pattern and Deep Convolutional Neural Networks, RP and Convolutional Neural Networks, it is found that the proposed RP+MobileNet-v3 model has the best performance in all aspects with diagnosis accuracy, parameter numbers, and Graphics Processing Unit usage, overcoming the over-fitting phenomenon and increasing the anti-noise performance. It is concluded that the proposed RP+MobileNet-v3 model has a higher diagnostic accuracy with fewer parameters and is a lighter model.

Diagnosing Faults in Power Transformers With Variational Autoencoder, Genetic Programming, and Neural Network

This work presents a new approach for the diagnosis of incipient faults in power transformers by considering dissolved gas analysis (DGA). A multilayer perceptron (MLP) neural network was trained to diagnose the type of transformer fault. For training and testing of the classifier, data were used from in-service transformers obtained from the IEC TC 10 database and other data obtained from the literature. To address the imbalance of the data from the database adopted and thus improve the generalization power of the classifier, a data augmentation technique based on a variational autoencoder neural network was used. For the selection and extraction of characteristics from the inputs to the classifier, a technique based on genetic programming (GP) is proposed, which allows the creation of a new n-dimensional space of characteristics, providing a greater ability to increase interclass distances and intraclass compaction. For the performance analysis of the proposed classifier, comparisons were made using the classification results obtained through the IEC 60599 conventional fault diagnosis method and other trained MLPs without the use of data augmentation and the proposed characteristics extractor. The results obtained demonstrate the applicability of the proposed methodology for fault diagnosis, with the proposed system obtaining an accuracy of 95.18% in the test basis, which is higher than the results achieved by the other methods used to perform a comparison and analysis of results.

Intelligent fault diagnosis algorithm of rolling bearing based on optimization algorithm fusion convolutional neural network.

As an essential component of mechanical equipment, the fault diagnosis of rolling bearings may not only guarantee the systematic operation of the equipment, but also minimize any financial losses caused by equipment shutdowns. Fault diagnosis algorithms based on convolutional neural networks (CNN) have been widely used. However, traditional CNNs have limited feature representation capabilities, thereby making it challenging to determine their hyperparameters. This paper proposes a fault diagnosis method that combines a 1D-CNN with an attention mechanism and hyperparameter optimization to overcome the aforementioned limitations; this method improves the search speed for optimal hyperparameters of CNN models, improves the diagnostic accuracy, and enhances the representation of fault feature information in CNNs. First, the 1D-CNN is improved by combining it with an attention mechanism to enhance the fault feature information. Second, a swarm intelligence algorithm based on Differential Evolution (DE) and Grey Wolf Optimization (GWO) is proposed, which not only improves the convergence accuracy, but also increases the search efficiency. Finally, the improved 1D-CNN alongside hyperparameters optimization are used to diagnose the faults of rolling bearings. By using the Case Western Reserve University (CWRU) and Jiangnan University (JNU) datasets, when compared to other common diagnosis models, the results demonstrate the usefulness and dependability of the DE-GWO-CNN algorithm in fault diagnosis applications by demonstrating the increased diagnostic accuracy and superior anti-noise capabilities of the proposed method. The fault diagnosis methodology presented in this paper can accurately identify faults and provide dependable fault classification, thereby assisting technicians in promptly resolving faults and minimizing equipment failures and operational instabilities.

Novel Bootstrap-Based Discriminant NPE Integrated With Orthogonal LPP for Fault Diagnosis

It has become a challenge to identify the discriminant information and the local geometric feature from the complex process data for improving fault diagnosis accuracy. Facing this challenge, this article proposes a novel fault diagnosis methodology based on discriminant neighborhood preserving embedding integrated with orthogonal locality preserving projections (DNPEOLPP). In DNPEOLPP, a novel local discriminant weight is constructed to identify the local geometric feature for dimensionality reduction (DR), so that the label of data is fully utilized. Furthermore, the cosine distance considering the spatial features of data is introduced into the process of neighbor graph construction to extract representative features. The bootstrap-based DNPEOLPP (B-DNPEOLPP) is eventually developed to further address the issue of singular matrix caused by the small number of classes when applying DNPEOLPP for DR. In addition, the Akaike information criterion (AIC) is utilized to identify the suitable order of DR, and AdaBoost.M2 is utilized to classify the type of fault. Finally, the Tennessee Eastman process (TEP) and the grid-connected photovoltaic system (GPVS) are used to validate the performance of B-DNPEOLPP. Through case study analysis, it is demonstrated that B-DNPEOLPP can effectively identify different faults and obtain superior accuracy in fault diagnosis.

An intelligent technique for fault detection and localization of three-level ANPC inverter with NP connection for electric vehicles

An intelligent technique for detecting and localizing an inverter switch fault or phase fault of a Three-Level Active Neutral Point Clamped (ANPC) inverter is proposed in this research. Moreover, a 3L-ANPC inverter can gain the controllability of EV's power train and not need to be stalled even after the occurrence of the fault. Hence, an efficient fault diagnosis methodology is required to identify the type of phase fault by a Support Vector Machine (SVM), a machine learning model consisting of sets of labeled training data with regression and classification challenges. Finally, when the fault occurs, the location of the switch fault can be identified by a Deep Neural Network (DNN), which consists of layers of neurons between the input and output layers which fuses the feature extraction process with increased accuracy. Thus, the detection and localization of the open-circuit fault of the switches in the ANPC inverter help overcome all single faults, hence gaining its current controllability without stopping the vehicle. The accuracy of fault detection is improved in a precise manner. Finally, the performance of the proposed work is evaluated over other conventional models concerning varied metrics like the accuracy of identification and localization.