Characteristics extraction and pattern recognition of partial discharge signals in high-voltage DC cable accessories
高压直流电缆附件局部放电信号的特征提取与模式识别

摘要

高压直流电缆在电力传输中地位关键，其附件局部放电信号分析关乎电力系统稳定。本研究围绕信号特征提取与模式识别展开，通过小波变换、经验模态分解等方法，从时域、频域和时频域精准提取信号特征；运用支持向量机、随机森林等机器学习算法，结合卷积神经网络、循环神经网络等深度学习模型，实现放电信号的准确识别。实验证实，该方法在复杂环境下识别效果良好，为电缆状态监测与故障预警提供有力支撑。

Abstract

High-voltage DC cables play a crucial role in power transmission, and the analysis of partial discharge signals in their accessories is vital for the stability of power systems. This study focuses on signal feature extraction and pattern recognition, employing methods such as wavelet transform and empirical mode decomposition to accurately extract features from the time domain, frequency domain, and time-frequency domain. By integrating machine learning algorithms like support vector machines and random forests with deep learning models such as convolutional neural networks and recurrent neural networks, the study achieves accurate identification of partial discharge signals. Experiments have shown that this method performs well in complex environments, providing robust support for cable condition monitoring and fault prediction.