Search

As a protective component of HV equipment, the primary function of a surge arrester is to mitigate the impact of surge voltages. When a surge arrester fails, the equipment it protects becomes vulnerable to damage. This study integrates chaotic systems with Convolutional Neural Networks (CNN) to diagnose faults in HV surge arresters. The Partial Discharge (PD) test was initially performed on six HV surge arrester fault models. The Discrete Wavelet Transform (DWT) was performed for filtering the PD signals. Subsequently, the Chen‐Lee chaotic system converted the filtered PD signals into a dynamic error scatter diagram, creating a feature map of various fault states. This feature map was then used as the input layer to train the CNN model. The results demonstrate that the proposed CNN achieved an accuracy of 97.0%, outperforming AlexNet and traditional methods using Histograms of Oriented Gradients (HOG) combined with Support Vector Machine (SVM), Decision Tree (DT), Backpropagation Neural Network (BPNN), and K‐Nearest Neighbor (KNN). This study also incorporates the LabVIEW graphic control software with a fault diagnosis system for HV surge arresters. The PD data can identify the fault type in real‐time, enhancing power equipment maintenance efficiency. [ABSTRACT FROM AUTHOR]

This study proposes an improved few-shot learning model of the Siamese network residual Visual Geometry Group (VGG). This model combined with time–frequency domain transformation techniques effectively enhances the performance of across-load fault diagnosis for induction motors with limited data conditions. The proposed residual VGG-based Siamese network consists of two primary components: the feature extraction network, which is the residual VGG, and the merged similarity layer. First, the residual VGG architecture utilizes residual learning to boost learning efficiency and mitigate the degradation problem typically associated with deep neural networks. The employment of smaller convolutional kernels substantially reduces the number of model parameters, expedites model convergence, and curtails overfitting. Second, the merged similarity layer incorporates multiple distance metrics for similarity measurement to enhance classification performance. For cross-domain fault diagnosis in induction motors, we developed experimental models representing four common types of faults. We measured the vibration signals from both healthy and faulty models under varying loads. We then applied the proposed model to evaluate and compare its effectiveness in cross-domain fault diagnosis against conventional AI models. Experimental results indicate that when the imbalance ratio reached 20:1, the average accuracy of the proposed residual VGG-based Siamese network for fault diagnosis across different loads was 98%, closely matching the accuracy of balanced and sufficient datasets, and significantly surpassing the diagnostic performance of other models. [ABSTRACT FROM AUTHOR]

As the proportion of renewable energy generation systems increases, traditional power generation facilities begin to face challenges, such as reduced output power and having the power turned off. The challenges are causing changes in the structure of the power system. Renewable energy sources, mainly wind and solar energy cannot provide stable inertia and frequency regulation capability. Ultimately, the power system's emergency response capability to face an N-1 is reduced, which leads to a reduction in system stability. Therefore, the application technology of the battery energy storage system is used to support the impact of changes in the new power system structure. This paper designed control technologies based on the WECC second-generation generic model, namely, dynamic regulation, steady regulation, and virtual inertia regulation. The models and control strategies are verified on Taiwan's 2025 power system target conditions, which consider the expected capacities for battery energy storage systems, and renewable energy sources with different load and N-1 fault levels. According to the simulation results, the capabilities of the RoCoF limitation, frequency nadir, frequency recovery, and system oscillation regulation are evaluated in the proposed strategies. Finally, the analysis results can help power operators make informed decisions when selecting and deploying battery energy storage systems. [ABSTRACT FROM AUTHOR]

This paper proposes an optimization classification model, which combines particle swarm optimization (PSO) with weighted knearest neighbors (WKNN), namely PWKNN. The model optimizes the weight and k parameter of WKNN to improve the detection accuracy of gearbox lubrication levels. In the experiment, the current signals of the generator are measured, and the relative frequency spectrum of the measured signals is illustrated by using fast Fourier transform (FFT). The features from the spectrum are extracted, and then the optimal weight and k parameter of WKNN are obtained by using PSO. The average detection accuracy of gearbox lubrication levels is 96% by using PWKNN, which the result shows that the proposed PWKNN can efficiently detect the lubrication level of gearboxes. The experiment also shows that the performance of the proposed PWKNN by using the current signals of the generator is superior to that by using typical vibration signals of a gearbox. In addition, the accuracy can reach 95.4% even in environments with 20 dB noise interference.

To accurately diagnose the XLPE power cable insulation fault, this research proposed a novel hybrid algorithm combined with Convolutional Probabilistic Neural Network (CPNN) based on Discrete Wavelet Transform (DWT) and Symmetrized Dot Pattern (SDP) analysis. First, it built seven different power cable insulation defect models to measure partial discharge signals of power cable insulation faults. Then, a discrete wavelet transform was used for noise filtering. The time‐domain partial discharge signal was directly converted into the point coordinate feature image of visual polar coordinates by SDP analyses. Finally, the feature image was trained and recognized by CPNN. After the important feature information of the feature‐image was extracted by convolution layer and pooling layer operations, it is applied to the power cable insulation fault state diagnosis system based on the rapid learning and highly parallel computing of Probabilistic Neural Network (PNN). The experimental results proved that the method proposed in this research could accurately diagnose the power cable insulation fault type and the recognition accuracy is higher than 96%. The proposed method has a short detection time and high diagnostic accuracy. This proves that it can be applied to detect the power cable insulation fault type. [ABSTRACT FROM AUTHOR]