基于多维量测数据序列的电压互感器故障识别方法

Fault Identification Method of Voltage Transformer Based on Multidimensional Measurement Data Sequence

摘要:

电压互感器量测数据是电力系统监测与控制的基础，错误的量测数据可能导致自动装置错误动作和调度人 员错误决策，因此，及时发现电压互感器故障至关重要。提出了一种基于多维量测数据序列的电压互感器故障识别 方法，该方法能够在线监测所有类型的电压互感器。首先，利用经验模态分解方法对电压互感器测量数据进行去噪 处理，消除坏数据干扰; 其次，采用时间序列分层聚类方法对多维电压互感器量测数据进行分析，识别出故障的电压 互感器; 最后，将所提方法应用于某实际案例，验证了其合理性和有效性。

Abstract:

The measurement data of voltage transformer is the basis of monitoring and control of power system，false measurement data may result in malfunction of automatic device and inaccurate decision of operators，therefore，identifying the fault of voltage transformer timely is quite important． A fault identification method for voltage transformer is proposed based on multidimensional measurement data sequence，which can monitor all kinds of voltage transformer on line． Firstly，the empirical mode decomposition method is used to denoise the measurement data of voltage transformer so as to eliminate the disturbance of bad data． Secondly，the time series hierarchical clustering method is adopted to analyze the multidimensional measurement data of voltage transformer so as to identify the faulted voltage transformer． Finally，the proposed method is applied to a practical case and is verified to be reasonable and valid．