Prioritizing preventive maintenance of the feeders, laterals, and MV substations of Semnan using the SOM

This paper presents a method for specifying the priority of the feeders, laterals, and medium voltage electrical substations for preventive maintenance (PM) plan throughout Semnan, Iran. For this purpose, the information set of ۲۰kV feeders from all the sub-transmission substations in the city along with ۶ features were gathered and were clustered into a number of clusters using clustering methods. In the end, the feeders with the most number of outages, longest repairing time and highest energy not supplied (ENS) were selected as the feeders with the highest priority for performing preventive maintenance. This method is also applicable to the laterals and distribution substations. Finally, better planning for preventive maintenance can be performed and leads to a significant decrease in the overall energy not supplied (ENS) using this method.

Keywords:

data mining , clustering , neural network , SOM