基于SOM的变压器故障诊断研究

通过变压器故障诊断能及时发现变压器的故障,以往应用比较广泛的故障分类方法是基于DGA结果的比值法,但其存在比较严重的缺陷。通过采用自组织映射网络对变压器故障进行诊断分类,结果表明自组织映射网络对变压器故障的分类准确且快速,能够降低检测人员的分析难度,提高诊断速度,对于现场变压器故障诊断应用前景广阔。

Transformer Fault Diagnosis Based on SOM

The fault of transformer can be found through power transformer fault diagnosis.The ratio method based on the results of the DGA is applied broadly,but there are serious defects.Self-Organization Map is used for diagnosis and classification of power transformer faults.The findings showed that self-organizing map network classifies power transformer fault accurately and quickly.It can reduce analysis difficulty and improve the speed of diagnosis for testing personnel.The application prospect of this method is broad for on-site fault diagnosis.