A Unified Framework With Incremental Learning Capacity for Industrial Fault Detection and Classification

Fast and reliable fault detection is critical in photovoltaic (PV) systems to improve reliability and energy yield and reduce maintenance costs, ensuring safe and efficient operation under varying operating conditions. Although recent data-driven PV fault detection techniques (FDTs) in literature have demonstrated high diagnostic accuracies, they often suffer from practical limitations, offline operation, lack of fault localization and/or inability to concurrently identify faults. To address these challenges, a unified framework is proposed that simultaneously integrates real-time operation, fault classification and localization, and concurrent-fault identification in a single compact diagnostic approach. This is realized by developing a parallel feedforward neural network (FFNN) architecture whose performance is enhanced by a residual model-based structure, resulting in a more interpretable, scalable, reliable and accurate scheme. In addition, Grey Wolf Optimizer–Support Vector Machine (GWO–SVM) feature selection is incorporated to select the most influential diagnostic features, thus reducing data redundancy, enhancing diagnostic accuracy, and shortening training time. The proposed approach was tested to diagnose five types of PV faults (open circuit, short circuit, partial shading, degradation, and simultaneous faults), as well as classify their intensity and location. Simulation results show that the proposed FFNNs consistently outperform conventional Support Vector Machines (SVMs) in classification accuracy, with accuracies exceeding 98% and 99% for fault classification and localization, respectively, and above 95% for noisy data. Moreover, GWO-SVM proved to offer more stable feature subsets compared to Particle Swarm Optimization–SVM (PSO–SVM) in the considered feature selection structure. Simulation results validated the effectiveness of the proposed unified multiclassification fault diagnosis approach, even under system uncertainties, making it suited for real-world PV systems.

Accurate fault classification and detection for the microgrid (MG) becomes a concern among the researchers from the state-of-art of fault diagnosis as it increases the chance to increase the transient response. The MG frequently experiences a number of shunt faults during the distribution of power from the generation end to user premises, which affects the system reliability, damages the load, and increases the fault line restoration cost. Therefore, a noise-immune and precise fault diagnosis model is required to perform the fast recovery of the unhealthy phases. This paper presents a review on the MG fault diagnosis techniques with their limitations and proposes a novel discrete-wavelet transform (DWT) based probabilistic generative model to explore the precise solution for fault diagnosis of MG. The proposed model is made of multiple layers with a restricted Boltzmann machine (RBM), which enables the model to make the probability reconstruction over its inputs. The individual RBM layer is trained with an unsupervised learning approach where an artificial neural network (ANN) algorithm tunes the model for minimizing the error between the true and predicted class. The effectiveness of the proposed model is studied by varying the input signal and sampling frequencies. A level of considered noise is added with the sample data to test the robustness of the studied model. Results prove that the proposed fault detection and classification model has the ability to perform the precise diagnosis of MG faults. A comparative study among the proposed, kernel extreme learning machine (KELM), multi KELM, and support vector machine (SVM) approaches is studied to confirm the robust superior performance of the proposed model.

Increasingly there is a need for fast, accurate, and sensitive detection of equipment and process faults to maintain high process yields and rapid fault classification (diagnosis) of the cause in order to minimize tool downtime in semiconductor manufacturing. Current methods treat fault detection and classification as a two-step process. We present a novel method to simultaneously detect and classify faults in a single-step using fault-specific control charts. These control charts are designed to discriminate between specific fault classes and the normal process operation as well as all other fault classes. Using a set of experimental data collected from an industrial plasma etcher, we demonstrate that, if the fault-specific charts are constructed using an orthogonal linear discriminant approach, they are effective in simultaneously detecting and classifying a given fault. We also demonstrate that this methodology has improved sensitivity for detection of faults when compared to other commonly used methods of fault detection. © 2003 The Electrochemical Society. All rights reserved.

Switching LDS-based approach for process fault detection and classification

With the development of the Industrial Internet of Things (IIoT) technology, fault diagnosis has emerged as a critical component of its operational reliability, and machine learning algorithms play a crucial role in fault diagnosis. To achieve better fault diagnosis results, it is necessary to have a sufficient number of fault samples participating in the training of the model. In actual industrial scenarios, it is often difficult to obtain fault samples, and there may even be situations where no fault samples exist. For scenarios without fault samples, accurately identifying the unknown faults of equipment is an issue that requires focused attention. This paper presents a method for the normal-sample-based mechanical equipment unknown fault detection. By leveraging the characteristics of the autoencoder network (AE) in deep learning for feature extraction and sample reconstruction, normal samples are used to train the AE network. Whether the input sample is abnormal is determined via the reconstruction error and a threshold value, achieving the goal of anomaly detection without relying on fault samples. In terms of input data, the frequency domain features of normal samples are used to train the AE network, which improves the training stability of the AE network model, reduces the network parameters, and saves the occupied memory space at the same time. Moreover, this paper further improves the network based on the traditional AE network by incorporating a convolutional neural network (CNN) and a long short-term memory network (LSTM). This enhances the ability of the AE network to extract the spatial and temporal features of the input data, further improving the network's ability to extract and recognize abnormal features. In the simulation part, through public datasets collected in factories, the advantages and practicality of this method compared with other algorithms in the detection of unknown faults are fully verified.

With the rise of growing privacy concerns and "data silos," federated learning is gaining traction in wind turbine fault diagnosis. Most existing methods unrealistically assume that fault classes remain static over time. However, wind turbine data are collected as dynamic streams. Existing methods necessitate the storage of all historical fault class data and require real-time model retraining using this data whenever new faults are introduced. The high demands for storage, computation, and bandwidth are impractical as new data keeps coming in. Additionally, when these methods are applied to diagnose new fault classes in dynamic data streams with limited resources and heterogeneity across wind farms, the model suffers from significant memory degradation. To address these challenges, a federated incremental collaborative fault diagnosis method for dynamic data streams across multiple wind farms is proposed. First, a new fault class detection method is presented to ensure when and where to introduce new fault classes. Second, a balance between the plasticity and stability of the fault diagnosis model at each wind farm is proposed to alleviate the fading memory problem. Third, a global model adaptive compensatory method is presented to address the fading memory issue of the aggregated model caused by heterogeneity. Finally, the proposed method was validated with data from three real-world wind farms in Hubei, Jiangsu, and Yunnan provinces, China. The results showed that it effectively mitigates fading memory issues and outperforms several state-of-the-art methods.

Fault Detection and Diagnosis (FDD) in complex industrial systems requires methods that can handle uncertain operating conditions, soft thresholds, evolving sensor behavior, and increasing volumes of heterogeneous data. Traditional model-based or rule-driven approaches offer interpretability but lack adaptability, while purely data-driven and Large Language Model (LLM)-based methods often struggle with consistency, traceability, and causal grounding. Dynamic Master Logic (DML) provides a causal and temporal reasoning structure with fuzzy rules that capture gradual drift, soft limits, and asynchronous sensor signals while preserving traceability and deterministic evidence propagation. Building on this foundation, this paper presents a DML-LLM hybrid architecture that integrates targeted LLM inference to interpret unstructured information such as logs, notes, or retrieved documents under controlled prompts that maintain domain constraints. The combined system integrates Bayesian updating, deterministic routing, and semantic interpretation into a unified FDD pipeline. In a semiconductor manufacturing case study, the proposed framework reduced time to detection (TTD) from 7.4 h to 1.2 h and improved the F1 score from 0.59 to 0.83 when compared with conventional Statistical Process Control (SPC) and Fault Detection and Classification (FDC) workflows. Provenance completeness increased from 18% to 96%, while engineer triage time was reduced from 72 min to 18 min per event. These results demonstrate that the hybrid framework provides a scalable and explainable approach to anomaly detection and fault diagnosis in sensor-rich industrial environments.

In this paper, we address the problem of fault classification in PhotoVoltaic (PV) arrays using a semi-supervised graph signal processing approach. Traditional fault detection and classification methods require large amounts of labeled data for training. In utility scale solar arrays, obtaining labeled data for different fault classes is resource intensive. We propose a graph based classification technique that relies on a limited amount of labeled data. We compare our results with the well known supervised machine learning classifiers such as the K-nearest neighbour classifier, random forest classifier, support vector machines, and artificial neural networks. We also show that the graph-based classifiers require lower training computational cost compared to the standard supervised machine learning algorithms. The proposed method also achieves good classification performance with unseen data. We validate our method on a real-time dataset and show significant improvements over existing approaches.

Acoustical damage detection of wind turbine blade using the improved incremental support vector data description

While detecting anomalies in hyperspectral imagery with support vector data description (SVDD), large numbers of operation was run because of the high dimension character of dataset and the complexity of background and the high miss rate was discovered because of the interfered background by interior anomalies. This paper used incremental support vector data description (ISVDD) method that samples are divided many sub-sample, and incremental study is designed to simplify the computation. On every sub-sample study, optimization is needed according of the support vectors obtained from above sub-sample and current sub-sample data. By the experiment on the HYMAP data, computation complexity of the algorithm decrease obviously and the computation speed increase highly under the similar detection effect compared with SVDD algorithm.

In real industrial processes, fault diagnosis methods are required to learn from limited fault samples since the procedures are mainly under normal conditions and the faults rarely occur. Although attention mechanisms have become increasingly popular for the task of fault diagnosis, the existing attention-based methods are still unsatisfying for the above practical applications. First, pure attention-based architectures like transformers need a substantial quantity of fault samples to offset the lack of inductive biases thus performing poorly under limited fault samples. Moreover, the poor fault classification dilemma further leads to the failure of the existing attention-based methods to identify the root causes. To develop a solution to the aforementioned problems, we innovatively propose a supervised contrastive convolutional attention mechanism (SCCAM) with ante-hoc interpretability, which solves the root cause analysis problem under limited fault samples for the first time. First, accurate classification results are obtained under limited fault samples. More specifically, we integrate the convolutional neural network (CNN) with attention mechanisms to provide strong intrinsic inductive biases of locality and spatial invariance, thereby strengthening the representational power under limited fault samples. In addition, we ulteriorly enhance the classification capability of the SCCAM method under limited fault samples by employing the supervised contrastive learning (SCL) loss. Second, a novel ante-hoc interpretable attention-based architecture is designed to directly obtain the root causes without expert knowledge. The convolutional block attention module (CBAM) is utilized to directly provide feature contributions behind each prediction thus achieving feature-level explanations. The proposed SCCAM method is testified on a continuous stirred tank heater (CSTH) and the Tennessee Eastman (TE) industrial process benchmark. Three common fault diagnosis scenarios are covered, including a balanced scenario for additional verification and two scenarios with limited fault samples (i.e., imbalanced scenario and long-tail scenario). The effectiveness of the presented SCCAM method is evidenced by the comprehensive results that show our method outperforms the state-of-the-art methods in terms of fault classification and root cause analysis.

A fault detection and diagnosis method for the hydraulic servo system based on adaptive threshold and self-organizing map (SOM) neural network is proposed in this study. The nonlinear, time-varying, fluid-solid coupling properties of the hydraulic servo system are considered. Fault detection is realized based on a two-stage radial basis function (RBF) neural network model. The first-stage RBF neural network is adopted as a fault observer for the hydraulic servo system; the residual error signal is generated by comparing the estimated observer output with the actual measurements. To overcome the drawback of false alarms when the traditional fixed fault threshold is used, an adaptive threshold producer is established by the second-stage RBF neural network. Fault occurrence is detected by comparing the residual error signal with the adaptive threshold. When a system fault is detected, the SOM neural network is employed to implement fault classification and isolation by analyzing the features of the residual error signal. Three types of common faults are simulated to verify the performance and effectiveness of the proposed method. Experimental results demonstrate that the proposed method based on adaptive threshold and SOM neural network is effective in detecting and isolating the failure of the hydraulic servo system.

A method of bearing fault diagnosis and classification based on Empirical Mode Decomposition (EMD) and Independent Component Analysis (ICA) is presented. There are aliasing phenomenon and correlation among intrinsic mode functions decomposed by EMD. Through elimination of information redundancy, an estimated intrinsic mode functions that includes much information of fault is separated for fault diagnosis and classification by the method of ICA, in fault diagnosis, multi-signals are composed of one engineering signal sampled by single sensor by using time-delay technique, and a fault sub-signal for fault diagnosis is separated from multi-signals by the method of EMD-ICA. In fault classification, based on fault sub-signal estimated by EMD-ICA, a parameter vector is composed of the coefficients of frequency, the residual of multi-information, correlation coefficients, entropy of frequency, approximate entropy and kurtosis, and this parameter vector is regarded as input of general regression neural network for judging bearing's three fault type.

Applications of Artificial Immune System (AIS) has been widely applied in various engineering, including network security, pattern recognition, combinational optimization, machine learning and fault diagnosis, etc. Fault diagnosis is another AIS application field directly mapped from the theory of immunity after information security and has made certain achievements in research. Real-valued Negative selection algorithms (RNSA) of AIS generate their detector sets based on the points of self data. Self data is regarded as the normal pattern of behavior of the monitored system. This paper provide a new fault detection method based on RNSA of artificial immunity. It can effectively overcome the deficiency of the various fault detection methods of today that cannot implement fault detections because there are only normal samples and not enough fault samples and short of the function of continuous learning. The test result shows that, by increasing a certain number of training samples, the accuracy of fault diagnosis has made great changes. This way has obvious advantage in robustness and accuracy in detection and shows a favorable prospect of application.

SummaryConventional fault recognition algorithms can only recognize the classes of faults that have appeared in wind turbine systems. However, if a new category of faults appears, the traditional algorithm can only misclassify it into the class of pre‐existing faults. In this paper, a new class fault recognition method based on deep learning is proposed. Firstly, the initialized model is built using known fault data, including detectors and classifiers. Secondly, the new class fault data detected by the detectors are put into the cache, and when the cache overflows, the new class faults are augmented with data using generative adversarial networks. Finally, the new class fault data and the generated data are added to the initial training set, and the structure and weights of the initialized model are updated using the new training set. In this way, the purpose of recognizing the new class of faults is achieved. The experimental results show that the proposed detection algorithm can effectively detect new class faults and the model can efficiently solve the problem of the new class of fault recognition after data augmentation.