基于随机矩阵理论的电子式互感器误差状态相关性分析方法

Error state correlation analysis based on random matrix theory for electronic transformer

现场复杂环境下电子式互感器误差影响量交叉作用,解耦困难,确定电子式互感器误差状态和各影响量之间的内在联系较为困难。为此,提出了一种基于高维随机矩阵理论的误差状态相关性分析方法。基于滑动时间窗实时构建了评估矩阵,基于卡尔曼滤波器实现了稀疏情况下随机矩阵的扩展,建立了相关性评价指标d_(MSR)和I_(MSR),明确了相关性分析步骤。采用所提方法对互感器误差状态监测平台的运行数据进行了算例分析,结果表明所提方法能够确定1个或多个影响量对电子式互感器误差状态的作用程度。

Since the influencing factors of electronic transformer error involve interactions between each other in com- plex field conditions, resulting in difficult decoupling implementation. Therefore, it is difficult to confirm the relation- ship between the error state of electronic transformer and the influencing factors. In order to solve this problem, an error state correlation analysis method for electronic transformer is proposed based on high-dimensional random matrix theory. The evaluation matrix is constructed in real-time using moving time-window and then extended based on Kalman filter in the case of sparse matrix. Two evaluation indexes of correlation level,i.e, dMsR and IMSR are pro- posed and the steps of correlation analysis are confirmed. The proposed method is adopted to analyze the operating data of the electronic transformer error state monitoring platform, the results show that it can be utilized to confirm the influencing degree of single or multiple influencing factors on the error state of electronic transformer.