基于混合分类器的高压断路器故障诊断  

Fault Diagnosis of High Voltage Circuit Breaker Based on Hybrid Classifier

作　　者：黄新波[1] 许艳辉 朱永灿[1] HUANG Xinbo;XU Yanhui;ZHU Yongcan(College of Electronics and Information,Xi’an Polytechnic University,Xi’an 710048,China)

机构地区：[1]西安工程大学电子信息学院,西安710048

出　　处：《高压电器》2022年第10期149-157,共9页High Voltage Apparatus

基　　金：国家自然科学基金资助项目(51707141);陕西省重点项目—工业领域(2018ZDXM-GY-040);西安市科技计划项目(201805030YD8CG14(4))。

摘　　要：为利用有限故障样本对高压断路器主要故障类型进行精准识别,文中提出了一种基于混合分类器的高压断路器故障诊断方法。首先采用改进F-Score特征选择算法进行特征选择,选择出合适的特征量子集作为构建混合分类器模型的输入参量。混合分类器分别由两个支持向量数据描述和混合粒子群算法优化的小波核函数孪生支持向量机共同组成去识别故障类型。两个并列的SVDD分别用来对正常或故障状态与已知故障或未知故障进行状态识别。HPSO-WTWSVM则用来准确识别已知故障类型。经高压断路器实例验证,表明所提新方法的分类精度优于其他传统方法。In order to make a fast and accurate judgment of the main fault types of high voltage circuit breakers with limited fault data samples, this paper proposes a hybrid classifier based fault diagnosis method for high voltage circuit breakers. Firstly, the improved F-Score feature selection algorithm is used to select features, and the appropriate feature set is selected as the input parameter for constructing hybrid classifier model. The hybrid classifier is composed of two support vector data descriptions and a hybrid particle swarm optimization wavelet kernel function twin support vector machine to identify faults. Two parallel of SVDD are used to identify the status of a normal or fault condition with a known fault or an unknown fault, respectively. HPSO-WTWSVM is used to accurately identify the type of known fault. The verification of the high voltage circuit breaker shows that the classification accuracy of the proposed new method is better than other traditional methods.

关 键 词：高压断路器 支持向量数据描述 混合粒子群算法 小波核函数孪生支持向量机 故障诊断 

分 类 号：TM561[电气工程—电器]