Designing an‎d planning the alarm an‎d the status of assets of smart power grids using machine learning

10/21/2026 12:00:00 AM

Considering the rapid expansion of smart power grids, the use of artificial intelligence an‎d machine learning methods in the analysis an‎d management of grid data has become increasingly important. In this context, due to the high costs of distribution network equipment an‎d the complexity of their maintenance an‎d repair processes, attention to inspections an‎d preventive maintenance is regarded as one of the main pillars for improving the efficiency an‎d reliability of the network. Timely implementation of these measures, in addition to reducing operational costs, leads to increased network reliability without the need for undesired outages. In this study, an intelligent warning system for assessing the health condition of transformers in the power distribution network was designed an‎d implemented using machine learning models. The aim of developing this system is to predict the probability of transformer failures an‎d, consequently, to optimize preventive maintenance planning an‎d asset management processes. The results of this research indicate that the Support Vector Machine (SVM) model, with an accuracy of 96%, achieved the best performance among the eva‎luated models. Furthermore, the most influential features affecting the health of distribution network transformers, in order of importance, include the total number of switching operations per year, the elapsed time since the last inspection (in months), an‎d the loading level. The application of such a system can be an effective step toward increasing efficiency, reducing unexpected outages, enhancing network operational stability, an‎d moving toward intelligent management of power distribution infrastructure.

هوش مصنوعي , شبكه هاي هوشمند برق , يادگيري ماشين , مديريت دارايي , تعميرات پيشگيرانه

artificial intelligence , Smart power grids , Machine Learning , Asset Management , preventive maintenance

Mohammad Iranparast

Dr. Meysam Farajollahi , Dr. Reza Dashti