Data needs for reliability assessment of distribution systems

In this paper, a fuzzy-adaptive model for prediction of failure rates of overhead power distribution lines based on physical and weather related features is presented. The input data used and the results obtained are discussed. Also, results on sensitivity of failure rates to input features are provided. The overall aim of this presentation is to outline a methodology for feature-based prediction of failure rates of distribution system components. Since this method heavily depends of the input data and their accuracy, need for such databases is stressed.