Automatic Fault Research Articles

TBINet: fabric defect detection based on a top-down and bottom-up inference network

A key aspect of textile production is the use of automatic defect detection methods for quality control, which is a critical part of the current fabric defect detection system. Benefiting from the explosive development of convolutional neural networks, it has been proved to be feasible to apply it to salient object detection model in fabric defect detection, but how to learn more robust features and better feature fusion for saliency are still complicated tasks. This article presents a novel saliency detection method for fabric defect detection based on both top-down and bottom-up saliency inference combined with two-way information. The top-down process is to infer high-level saliency by gradually using higher-level and richer semantic features on the basis of the backbone. Then a self-fusion enhanced representation module is proposed, which simulates the parallel processing mechanism with residual connection in a top-down manner to generate robust features and effectively characterize the texture features of complex fabrics. In addition, in the down-top process, the interactive feature fusion module is designed, which is used for coarse-to-fine saliency estimation, where the high-level saliency is gradually integrated with finer lower-level features to obtain a fine-grained result. Finally, fabric defect detection is localized by segmenting the generated saliency map. Extensive experimental results on two kinds of fabric image datasets have demonstrated that the proposed TBINet can localize the defect region with high accuracy, and perform favorably against seven state-of-the-art methods on eight widely tested metrics.

An automatic rotor bar fault diagnosis using fuzzy logic and DWT-energy for backstepping control driven induction motor in low-speed operation

An automatic rotor bar fault diagnosis using fuzzy logic and DWT-energy for backstepping control driven induction motor in low-speed operation

Algorithm analysis of defect depth for PVC plate with flat-bottomed holes in active pulsed infrared thermography

Active infrared thermography (IRT) has been extensively employed in non-destructive testing in a wide variety of fields. It is capable of extracting defect information of tested object based on the infrared thermal image sequence. However, conventional infrared thermal images are often subjected to defect information with low pixel resolution, and defects are difficult to quantitatively analyze. By exploiting flat-bottomed holes in a PVC plate as defect specimens, a method for quantitative defect depth recognition based on the fusion principal component analysis (PCA) algorithm with sliding-window mechanism and the one dimensional—residual neural network—convolutional block attention module (1DResnet50_CBAM) model was proposed for the reconstructed image sequence of active IRT to address the above-described issues in this study. First, defect information and location were extracted from the original infrared sequence thermal image of the specimen using PCA algorithm with sliding-window mechanism. Then, the dimension of the defect data was reduced using the temporal characteristic of the infrared temperature field. That is, the three-dimensional defect data were transformed into 1D temporal infrared thermal signal. Moreover, the 1D infrared signal time series corresponding to the defect pixel points in the infrared sequence image served as the input to the network, and the defect depth served as the output for automatic defect detection and depth quantification. As indicated by the results, the proposed method based on the fusion PCA algorithm with sliding-window mechanism and 1DResnet50_CBAM model is capable of accurately detecting and quantifying defects. Compared with conventional prediction algorithms, the proposed model can more effectively extract defect information from the infrared detection images, with the defect depth relative prediction error less than 1.5%. Thus, the proposed model was confirmed as an effective method and model for defect recognition and quantitative analysis using infrared thermal detection technology.

LSA-Net: Location and shape attention network for automatic surface defect segmentation

Neural network algorithms for segmenting defects are widely used in industrial production. However, how to fuse the location information of defects in a single model and avoid the features extracted by different submodules gradually tend to be similar during the training is still a problem. To solve these problems, an end-to-end network is proposed that focuses on defect location and shape features, which can guarantee the difference between features extracted by different submodules. In the encoding stage, the location attention module enhances the perception of defect locations. The shape detection module with feature difference loss is designed to strengthen the detection of defect shapes. In the decoding stage, the features of different scales are fused to obtain the final defect region. The experimental results confirm the effectiveness of the proposed location and shape detection modules in the intersection over union on four datasets.

A Definition Rule for Defect Classification and Grading of Solar Cells Photoluminescence Feature Images and Estimation of CNN-Based Automatic Defect Detection Method

A nondestructive detection method that combines convolutional neural network (CNN) and photoluminescence (PL) imaging was proposed for the multi-classification and multi-grading of defects during the fabrication process of silicon solar cells. In this paper, the PL was applied to collect the images of the defects of solar cells, and an image pre-processing method was introduced for enhancing the features of the defect images. Simultaneously, the defects were defined by 13 categories and three divided grades of each under the definition rules of defects that were proposed in accordance with distribution and characteristics of each defect category, and expand data were processed by various data augmentation. The model was therefore improved and optimized based on the YOLOv5 as the feature extractor and classifier. The capability of the model on distinguishing categories and grades of solar cell defects was improved via parameter tuning and image pre-processing. Through experimental analysis, the optimal combination of hyperparameters and the actual effect of data sample pre-processing on the training results of the neural network were determined. Conclusively, the reasons for the poor recognition results of the small target defects and complex feature defects by the current model were found and further work was confirmed under the foundation of the differences in recognition results between different categories and grades.

Aeroengine Blade Surface Defect Detection System Based on Improved Faster RCNN

Aiming at the difficulty of automatic blade detection and the discontinuous defects on the full image, an aeroengine blade surface defect detection system based on improved faster RCNN is designed. Firstly, a dataset of blade surface defects is constructed. To solve the problem that the original faster RCNN is hard to detect tiny defects, RoI align is adopted to replace RoI pooling in the improved faster RCNN and the feature pyramid networks (FPN) combined with ResNet-50 are introduced for feature extraction. To address the issue of discontinuous defects on the full image, the nonmaximum suppression (NMS) algorithm is improved to ensure the continuity of defects. A four-degree-of-freedom (4-DOF) motion platform and an industrial camera are used to collect images of blade surfaces. The detection results generated by the improved faster RCNN are compared with the results of the unimproved method. The experimental results prove that the defect detection system based on the improved faster RCNN can realize automatic defect detection on the blade surface with high accuracy. It also solves the issues of tiny defect detection and discontinuous defects on the full result image of the blade.

A Momentum Contrastive Learning Framework for Low-Data Wafer Defect Classification in Semiconductor Manufacturing

Wafer bin maps (WBMs) are essential test data in semiconductor manufacturing. WBM defect classification can provide critical information for the improvement of manufacturing processes and yield. Although deep-learning-based automatic defect classification models have demonstrated promising results in recent years, they require a substantial amount of labeled data for training, and manual labeling is time-consuming. Such limitations impede the practical application of existing algorithms. This study introduces a low-data defect classification algorithm based on contrastive learning. By employing momentum contrastive learning, the network extracts effective representations from large-scale unlabeled WBMs. Subsequently, a prototypical network is utilized for fine-tuning with only a minimal amount of labeled data to achieve low-data classification. Experimental results reveal that the momentum contrastive learning method improves the defect classification performance by learning feature representation from large-scale unlabeled data. The proposed method attains satisfactory classification accuracy using a limited amount of labeled data and surpasses other comparative methods in performance. This approach allows for the exploitation of information derived from large-scale unlabeled data, significantly reducing the reliance on labeled data.

Optical Imaging Deformation Inspection and Quality Level Determination of Multifocal Glasses.

Multifocal glasses are a new type of lens that can fit both nearsighted and farsighted vision on the same lens. This property allows the glass to have various curvatures in distinct regions within the glass during the grinding process. However, when the curvature varies irregularly, the glass is prone to optical deformation during imaging. Most of the previous studies on imaging deformation focus on the deformation correction of optical lenses. Consequently, this research uses an automatic deformation defect detection system for multifocal glasses to replace professional assessors. To quantify the grade of deformation of curved multifocal glasses, we first digitally imaged a pattern of concentric circles through a test glass to generate an imaged image of the glass. Second, we preprocess the image to enhance the clarity of the concentric circles' appearance. A centroid-radius model is used to represent the form variation properties of every circle in the processed image. Third, the deviation of the centroid radius for detecting deformation defects is found by a slight deviation control scheme, and we gain a difference image indicating the detected deformed regions after comparing it with the norm pattern. Fourth, based on the deformation measure and occurrence location of multifocal glasses, we build fuzzy membership functions and inference regulations to quantify the deformation's severity. Finally, a mixed model incorporating a network-based fuzzy inference and a genetic algorithm is applied to determine a quality grade for the deformation severity of detected defects. Testing outcomes show that the proposed methods attain a 94% accuracy rate of the quality levels for deformation severity, an 81% recall rate of deformation defects, and an 11% false positive rate for multifocal glass detection. This research contributes solutions to the problems of imaging deformation inspection and provides computer-aided systems for determining quality levels that meet the demands of inspection and quality control.

An Ultralow-Loss Fault-Current-Limiting Z-Source LVDC Circuit Breaker Using Cryogenic Power Electronics

Room-temperature power electronics based DC circuit breakers and fault current limiters face two major challenges: high conduction loss, and time-delayed response. This letter presents a novel circuitry, for the first time (based on literature studies), to combine cryogenic power electronics (CPE), together with the superconducting Z-source DC circuit breaker technology. The design, fabrication and tests of the new device were carried out, and 3 advantages were proved: (1) The actual overall system loss (including cryogenic issues) reduced by 62.1% in the 2 kW case study; (2) The no-delay automatic fault current limiting technology provided a sudden protective resistor ∼0.1 Ω to suppress the fault below 200 A, and meanwhile stabilized the bus voltage; (3) The CPE-based breaker achieved both the active device isolation and passive fault isolation within 60 μs.

A Power Grid Fault Diagnosis Method Considering Online Dynamic Alarm Information

With the continuous development of the smart grid and the gradual improvement of the demand for grid reliability and power grid self-healing capability, a large number of artificial intelligence technologies have been applied to the power grid. In order to improve the automation of the power grid, realize automatic fault identification and diagnosis after a power grid failure, and improve the efficiency of fault handling by the dispatching department, it is of great practical significance to use alarm information to achieve accurate fault judgment. However, the difference between the dynamic reading of alarm information flow and static samples in engineering applications is often ignored in the research of fault diagnosis using alarm information, which makes it difficult to apply fault diagnosis methods in engineering practice. Based on the characteristics of the LSTM algorithm, such as deep knowledge extraction ability, data mining ability and time series knowledge processing ability, this paper studies the power grid fault diagnosis method based on LSTM to adapt to the data of each sampling time window in the process of online fault diagnosis, solves the problems faced by the application of traditional diagnosis methods in engineering practice, and effectively improves the reliability of fault diagnosis.

Automatic detection surface defects based on convolutional neural networks and deflectometry

AbstractSurface defects in industrial refrigerator manufacturing processes can cause significant production losses and compromise product quality. This area is underexplored and currently, visual quality inspection is a subjective process that requires expert intervention, which limits process efficiency and can lead to errors in defect detection. This paper presents a novel approach for automatic surface defect detection using a combination of convolutional neural networks (CNN) and deflectometry. The proposed method takes advantage of the high accuracy and robustness of CNNs in image classification tasks and the sensitivity of deflectometry to detect subtle surface variations. First, a prototype was built to get the images from the refrigerator. Second, using video recordings, we captured surface topographic data using deflectometry, which we then use to generate surface images. Next, we train a CNN to classify the surface images as defective or normal. The proposed method offers a promising solution for automatic detection and quality control of surface defects in refrigerator manufacturing processes. However, this method could also improve the production of vehicles, household appliances in general, and any product that can suffer scratches and dents.

A CNN-Architecture-Based Photovoltaic Cell Fault Classification Method Using Thermographic Images

Photovoltaic (PV) cells are a major part of solar power stations, and the inevitable faults of a cell affect its work efficiency and the safety of the power station. During manufacturing and service, it is necessary to carry out fault detection and classification. A convolutional-neural-network (CNN)-architecture-based PV cell fault classification method is proposed and trained on an infrared image data set. In order to overcome the problem of the original dataset’s scarcity, an offline data augmentation method is adopted to improve the generalization ability of the network. During the experiment, the effectiveness of the proposed model is evaluated by quantifying the obtained results with four deep learning models through evaluation indicators. The fault classification accuracy of the CNN model proposed here has been drawn by the experiment and reaches 97.42%, and it is superior to that of the models of AlexNet, VGG 16, ResNet 18 and existing models. In addition, the proposed model has faster calculation, prediction speed and the highest accuracy. This method can well-identify and classify PV cell faults and has high application potential in automatic fault identification and classification.

Optimal periodicity-enhanced group sparse for bearing incipient fault feature extraction

Efficient and automatic fault feature extraction of rotating machinery, especially for incipient faults is a challenging task of great significance. In this article, an optimal periodicity-enhanced group sparse method is proposed. Firstly, a period sequence determination method without any prior information is proposed, and the amplitude is calculated by the numerical characteristics of the vibration signal to obtain period square waves. Secondly, the periodic square waves are embedded into the group sparse algorithm, to eliminate the influence of random impulses, and intensify the periodicity of the acquisition signal. Thirdly, a fault feature indicator reflecting both signal periodicity and sparsity within and across groups is proposed as the fitness of the marine predator algorithm for parameter automatic selection. In addition, the method proposed is evaluated and compared by simulation and experiment. The results show that it can effectively extract incipient fault features and outperforms traditional overlapping group shrinkage and Fast Kurtogram.

Application of a wavelet neural network approach to detect stator winding short circuits in asynchronous machines

Introduction. Nowadays, fault diagnosis of induction machines plays an important role in industrial fields. In this paper, Artificial Neural Network (ANN) model has been proposed for automatic fault diagnosis of an induction machine. The aim of this research study is to design a neural network model that allows generating a large database. This database can cover maximum possible of the stator faults. The fault considered in this study take into account a short circuit with large variations in the machine load. Moreover, the objective is to automate the diagnosis algorithm by using ANN classifier. Method. The database used for the ANN is based on indicators which are obtained from wavelet analysis of the machine stator current of one phase. The developed neural model allows to taking in consideration imbalances which are generated by short circuits in the machine stator. The implemented mathematical model in the expert system is based on a three-phase model. The mathematical parameters considered in this model are calculated online. The characteristic vector of the ANN model is formed by decomposition of stator current signal using wavelet discrete technique. Obtained results show that this technique allows to ensure more detection with clear evaluation of turn number in short circuit. Also, the developed expert system for the taken configurations is characterized by high precision.

Performance degradation of air source heat pumps under faulty conditions

The ongoing European regulations towards the decarbonization of energy intensive sectors such as heating and cooling will lead to an increase of heat pump appliances. Apart from anomalies that can be detected in a preliminary phase after the start-up, soft faults such as refrigerant leakage and heat exchanger fouling may cause a performance degradation in such systems that evolves in time and that must be detected before it becomes too big. From this perspective, in a previous paper (Pelella et al. [1]) it has been demonstrated that, in cooling mode, a seasonal performance penalization up to 50% can be reached in case of a not-planned maintenance scenario, whereas wherever a timed or intelligent maintenance strategy is implemented based on data-monitoring, this impact can be significantly reduced. Therefore, in order to extend the analysis previously carried out to the heating cases, and to evaluate the consequent energetic degradation of the system in case of soft faults occurring, depending on different climate conditions and maintenance scenarios, this paper develops a physic-based model to simulate system components and to describe the fault phenomenology on a residential 2.6 kW air-to-air heat pump, operating in winter mode. The results show that in heating mode a 40% condenser fouling and a 30% refrigerant leakage cause a performance degradation of respectively 16% and 12%, whereas in case of evaporator fouling the performance penalization is only of 3.2%. Moreover, the performance degradation is enhanced by the overlapping effect of simultaneous faults. Finally, from seasonal simulations of the heat pump along an entire machine lifetime of 12 years, it is found that none of the maintenance strategies analysed is able to significantly reduce the number of scenarios penalized by faults, opening to the potential development of systems for the automatic fault detection, diagnosis, and evaluation (FDDE).

Automatic Detection Method of Sewer Pipe Defects Using Deep Learning Techniques

Regular inspection of sewer pipes can detect serious defects in time, which is significant to ensure the healthy operation of sewer systems and urban safety. Currently, the widely used closed-circuit television (CCTV) inspection system relies mainly on manual assessment, which is labor intensive and inefficient. Therefore, it is urgent to develop an efficient and accurate automatic defect detection method. In this paper, an improved method based on YOLOv4 is proposed for the detection of sewer defects. A significant improvement of this method is using the spatial pyramid pooling (SPP) module to expand the receptive field and improve the ability of the model to fuse context features in different receptive fields. Meanwhile, the influence of three bounding box loss functions on model performance are compared based on their processing speed and detection accuracy, and the effectiveness of the combination of DIoU loss function and SPP module is verified. In addition, to address the lack of datasets for sewer defect detection, a dataset that contains 2700 images and 4 types of defects was created, which provides useful help for the application of computer vision techniques in this field. Experimental results show that, compared with the YOLOv4 model, the mean average precision (mAP) of the improved model for sewer defect detection are improved by 4.6%, the mAP can reach 92.3% and the recall can reach 89.0%. The improved model can effectively improve the detection and classification accuracy of sewer defects, and has significant advantages compared with other methods.

Automatic Fault Detection in Industrial Smart Grids Using KNN and Ensemble Classifiers

The use of sensitive electrical gadgets in industries, buildings, smart cities, and homes has increased drastically in recent years. PQ events such as interruptions, surges, and sags have a high impact on these sensitive devices. The failure of these delicate devices in real-time applications, particularly smart applications, may result in significant damage. The supply quality decreases because of the failure of internal transmission system elements, unbalanced loads, and other outdoor issues such as like weather. Several academics have proposed techniques to analyze these PQ disturbances, including wavelet packets, S-transform, rough sets and neural networks. In all the available algorithms, the classification procedure involves the extraction of a large set of features from the transformed outputs, training the classifier, and finally making a conclusion with the classifier. Because of the involvement of a large number of features, the computational cost of all these methods increases. To reduce complexity and enhance classification efficiency, the proposed method focuses on extracting fewer low-complexity wavelet features from signals. Pattern recognition (PR) methods, such as the wide variety of K-nearest neighbors (KNN) and ensemble classifiers, are used to classify PQ events in this study. The performance of the proposed ML approaches' performance is evaluated at various training and testing rates. Subsequently, the performance of the proposed strategies was compared to that of the current methods to determine the dominance of the proposed approaches.

Normal-Knowledge-Based Pavement Defect Segmentation Using Relevance-Aware and Cross-Reasoning Mechanisms

Automatic pavement defect segmentation is a big challenge because of the class diversity and extremely random distribution of defects. Most existing approaches focus on supervised strategies to achieve decent performance. Due to the difficulty of getting massive densely annotated samples and the limited prior knowledge of potential defects, these methods have significant bottlenecks in the actual pavement settings. This paper proposes a relevance-aware and cross-reasoning network (RCN) for anomaly segmentation of pavement defects, which can segment defects using merely non-defective images for training. A relevance-aware transformer-based encoder is first devised to model intrinsic interdependencies across local features, thus improving representations of complex non-defective images. Next, a dual decoder strategy is proposed to remap the encoder-generated latent dependencies at the local semantic and global detailed levels, respectively. Specifically, a cross-reasoning refinement module is built in the local decoder to reason the cross-relationship between spatial and channel dimensions. Finally, a context-aware abnormal distillation measurement is developed to evaluate the semantic reconstruction deviations during the inference. Under the guidance of semantic affinity, this measurement allows our model to highlight defective areas adaptively. Extensive experimental results on four datasets indicate that RCN outperforms other leading anomaly segmentation methods.

Automated Bone Fracture Detection Using Convolutional Neural Network

Bones are an important part of the human body. Humans are prone to bone fractures, which can occur as a result of tremendous pressure being applied to the bone, or as a result of a simple accident. As a result, in the medical field, an accurate diagnosis of a bone fracture is crucial. In this work, bone fractures are examined using X-ray/CT scans. The purpose of this research is to develop an image processing based system that can classify bone fractures rapidly and accurately using data from x-ray and CT images. In many diagnostic and therapeutic applications, automatic fault detection in MRI and CT images is critical. Tumor segmentation and classification are difficult due to the large amount of data in MR images and the fuzzy boundaries. This project proposed an automatic bone fracture detection system that improves accuracy and yield while cutting down on diagnosis time. The aim is to classify into two categories: normal, fractured. The amount of data in MR and CT pictures is too great for human interpretation and analysis. Bone fracture segmentation in Magnetic Resonance Imaging (MRI) has appeared as an emerging subject study in the realm of medical imaging systems in recent years. The potential to accurately recognise the size and location of a bone fracture is critical in the diagnosis of a fracture. Pre-processing of MR images, feature extraction, and classification are the four stages of the diagnosis procedure. The features are extracted using wavelet transformation after the image has been preprocessed (DWT). Automated fracture identification is essential in a computer-aided telemedicine system. Human arbitrary bones frequently fracture as a consequence of accidental traumas like slipping. Computer-aided diagnosis (CAD) relieves doctors’ workload while also detecting fractures. A new classification network sensitive to fracture lines is described, called Crack-Sensitive Convolutional Neural Network (CrackNet). Faster Region with Convolutional Neural Network (Faster R-CNN) is used to identify 20 various types of bone areas in X-ray pictures, and CrackNet is used to assess incase each bone area is broken. The classification accuracy for 100 training and test sets is 99.5%.

Deep learning based online metallic surface defect detection method for wire and arc additive manufacturing

Wire and arc additive manufacturing (WAAM) is an emerging manufacturing technology that is widely used in different manufacturing industries. To achieve fully automated production, WAAM requires a dependable, efficient, and automatic defect detection system. Although machine learning is dominant in the object detection domain, classic algorithms have defect detection difficulty in WAAM due to complex defect types and noisy detection environments. This paper presents a deep learning-based novel automatic defect detection solution, you only look once (YOLO)-attention, based on YOLOv4, which achieves both fast and accurate defect detection for WAAM. YOLO-attention makes improvements on three existing object detection models: the channel-wise attention mechanism, multiple spatial pyramid pooling, and exponential moving average. The evaluation on the WAAM defect dataset shows that our model obtains a 94.5 mean average precision (mAP) with at least 42 frames per second. This method has been applied to additive manufacturing of single-pass, multi-pass deposition and parts. It demonstrates its feasibility in practical industrial applications and has potential as a vision-based methodology that can be implemented in real-time defect detection systems.