主变压器解体检查试验中低压绕组故障诊断方法  

Fault Diagnosis Method for Low Voltage Winding in Main Transformer Disassembly Inspection Test

作　　者：虞晓巍 YU Xiaowei(Shanghai Jiulong Electric Power(Group)Co.,Ltd.,Transformer Repair and Testing Branch,Shanghai 200436,China)

机构地区：[1]上海久隆电力(集团)有限公司变压器修试分公司,上海200436

出　　处：《通信电源技术》2023年第23期226-228,共3页Telecom Power Technology

摘　　要：常规的低压绕组故障诊断方法,不能通过人工智能图像识别技术确定故障位置,致使故障诊断结果不准确,存在误差。因此,提出主变压器解体检查试验中低压绕组故障诊断方法。通过计算低压绕组短路电流,方便判断故障的大致位置,采用人工智能图像识别技术对于低压绕组故障位置进行详细定位,通过振荡波信号建立变压器低压绕组故障模型,然后借由图像颜色特征,在模型中提取小波包时频图信息,完成故障诊断。对故障点进行了对比实验,实验组的诊断准确率为100.0%,证明解体检查试验中低压绕组故障诊断方法具有有效性。Conventional fault diagnosis methods of low-voltage windings can't determine the fault location through artificial intelligence image recognition technology,which leads to inaccurate fault diagnosis results and errors.Therefore,a method for diagnosing low voltage winding faults in the disassembly inspection test of the main transformer is proposed.By calculating the short circuit current of the low voltage winding,it is convenient to determine the approximate location of the fault.Artificial intelligence image recognition technology is used to locate the fault location of the low voltage winding in detail.A transformer low voltage winding fault model is established through oscillation wave signals.Then,using the color features of the image,wavelet packet time-frequency map information is extracted from the model to complete the fault diagnosis.The fault points are compared in experiments,and the diagnostic accuracy of the experimental group is 100.0%,which proves that the fault diagnosis method of low-voltage winding in disassembly inspection test is effective.

关 键 词：变压器解体检查 低压绕组 主变压器 故障诊断 

分 类 号：TM407[电气工程—电器]