基于多传感信息融合的电力变压器绕组故障诊断方法

Multisensor-information-fusion-based Fault Diagnosis Transformer Windings

传统电力变压器绕组故障诊断方法直接对变压器绕组故障进行诊断,未对振动信号进行特征提取,造成诊断精准度低。为此,提出基于多传感信息融合的电力变压器绕组故障诊断方法。在故障诊断之前采集绕组振动信号,在采集绕组振动信号的基础上,重点对变压器绕组振动特征进行提取,通过提取的特征利用多传感信息融合算法对变压器绕组变形故障进行故障诊断。设计对比实验,实验结果证明该方法在不同数据预处理下的故障诊断精准度高于传统方法,有一定的研究价值。

Conventional methods of transformer winding fault diagnosis directly determine transformer winding faults without feature extraction of vibration signals,resulting in low diagnosis accuracy.Thereby the present study proposed a multisensor-information-fusion-based method to diagnose transformer winding faults.The method obtains winding vibration characteristics by extracting from winding vibration signals measured before performing diagnosis,and utilizes the extracted vibration characteristics and employs the multisensor information fusion algorithm to determine winding deformation faults.The proposed method was proved by comparative experiments to have superior diagnosis accuracy under different data preprocessing compared with conventional methods,and thereby,a certain research value.