This paper presents the methodology to predict the strong wind-related damage in the distribution system. For this, weather data and distribution system failure data are collected from the Meteorological Administration and an electric power company in South Korea respectively. Then, these datasets are integrated to analyze a relationship between weather and equipment failures in the distribution system. The prediction model is developed based on a typical regression model using the highly-correlated weather parameters selected by Pearson correlation analysis. Furthermore, two models which called the divided and combined models are developed to enhance an accuracy of prediction based on the natural disaster standard. To verify the algorithm, the results of the developed model are compared with the basic regression model by using the actual data in the strong windy day.
ACKNOWLEDGMENT This work was supported by the Korea Institute of Energy Technology Evaluation and Planning (KETEP) and the Ministry of Trade, Industry & Energy (MOTIE) of the Republic of Korea (No. 20172410100040).