Monitoring and identification of metal-oxide surge arrester conditions using multi-layer support vector machine

Metal-oxide surge arresters (MOSAs) are essential equipments for power system protection and devices from lightning and switching transient overvoltages. Therefore, their operating condition and diagnosis are very important. In this study, a multi-layer support vector machine (SVM) classifier has been used for MOSA conditions monitoring based on experimental tests. Three features are extracted based on the test results for determining surge arresters operating conditions including clean virgin, ultraviolet (UV) aged clean surface, surface contaminations after and before UV housing ageing, and degraded varistors along active column. Then, the multi-layer SVM classifier is trained with the training samples, which are extracted by the above data processing. Finally, the five fault types of surge arresters are identified by this classifier. The test results show that the classifier has an excellent performance on training speed and reliability which confirm the high applicability of introduced features for correct diagnostic of surge arresters conditions.