基于主成分分析和支持向量机的高压断路器机械状态识别方法

Mechanical Status Identification of High Voltage Circuit Breakers Based on Principal Component Analysis and Support Vector Machines

高压断路器分合闸过程中的触头行程曲线蕴含着反映其内部机构机械状态的丰富信息,是实现其状态识别和故障诊断的重要依据。文中提出了一种基于主成分分析和支持向量机的高压断路器机械状态识别方法,基于奇异值分解计算特征量的主成分,降低特征量的维度,筛选出包含主要信息的特征矩阵,然后基于特征矩阵构建支持向量机,结合交叉验证网格搜索确定最优参数,进而确定最优分类模型。对实验数据的分析结果表明,该方法可以有效提取触头行程曲线中蕴含的特征信息,准确度高达99%,可以实现对高压断路器机械状态的识别。

The contact travel curve of high voltage circuit breaker not only implies the valuable information of operation process that indicates the mechanical status of high voltage circuit breakers,but also is an important basis for its status recognition and fault diagnosis.In this paper,a mechanical status identification method based on principal component analysis and support vector machines is proposed for high voltage circuit breakers.Firstly,by calculating the principal component of each feature and reducing the dimension of the features based on singular value decomposition,the feature matrix containing the main information is selected.Secondly,some support vector machines are built based on feature matrix and the optimal parameters determined by cross-validation grid search to identify the mechanical status of high voltage circuit breakers.It is proved that this method is effective to extract the feature parameters from the contact travel curve by analyzing the experimental data.The support vector machines classifier with optimal feature matrix is 99% accurate,which can be used to classify the mechanical status of high voltage circuit breakers.