油纸绝缘匝间放电超声信号的发展特性和放电阶段识别方法

Development Characteristics of Ultrasonic Signals from Inter-Turn Discharge in Oil-Paper Insulation and Identification Methods for Discharge Stages

中文关键词:匝间放电  超声信号  阶段识别  斑翠鸟优化

keywords:inter-turn discharge  ultrasonic signal  stage identification  pied kingfisher optimization （PKO） algorithm

中文摘要:为了研究变压器油纸系统匝间局部放电，该研究构建了基于超声信号的检测试验平台，并采用恒压法获取数据。通过设计数据预处理方法、提取超声信号的时域和频域特征，分析特征参量的发展特性。基于发展特性将匝间局部放电划分为3个阶段，起始阶段、发展阶段和预击穿阶段，通过层次聚类算法验证其合理性，并构建了基于斑翠鸟优化算法的最小二乘支持向量机-自适应增强模型的阶段识别方法。实验结果表明，匝间超声信号特征在发展阶段波动较小，起始阶段和预击穿阶段有明显突变；3个阶段的匝间放电数据样本的识别正确率均超过96%，总体识别正确率为98%。

Abstract:In this study，a test platform for detecting inter-turn partial discharge （IPD） in transformer oil-paper insulation systems，based on ultrasonic signals，is established. Data acquisition is performed using the constant voltage method. Through designing a data preprocessing method for ultrasonic signals，the time-domain and frequency-domain characteristics of the signals are extracted，and the development characteristics of the characteristic parameters are analyzed. According to these evolutionary patterns，inter-turn partial discharge is categorized into three stages： the initial stage，the development stage，and the pre-breakdown stage. The hierarchical clustering algorithm is employed to validate the rationality of this staging. Additionally，a least squares support vector machine-adaptive boosting model，optimized by the pied kingfisher optimization algorithm，is constructed. Test results indicate that the characteristics of inter-turn ultrasonic signals exhibit minimal fluctuation during the development stage，whereas significant mutations occur in the initial stage and pre-breakdown stage. The recognition accuracy for inter-turn discharge data samples across the three stages exceeds 96%，with an overall recognition accuracy of 98%.