Reliability indices of a radial distribution system with sectionalizing as a function of network structure parameters

Since reliability indices are very important in customer services, they should be considered when a distribution system is designed. A method of estimating the average time of interruptions and the annual unsupplied energy caused by line failures is proposed in this paper. Radial single-feeder networks, typical for rural regions, are considered. It is shown that the most influential factors in determining system reliability are the area of the region, the angle of the feeder service zone, and the number of customers (load points). In particular, the effect of automatic sectionalizer implementation on unsupplied energy is evaluated using an optimal allocation algorithm which adopts the genetic search technique. The efficiency of the system sectionalizing method is estimated as a function of network structure parameters. A simulation study consisting of extensive computational experimentation was carried out to obtain a data set used to train neural networks. The weight coefficients and the structure of these networks are presented, so that functional dependences can be reproduced using available neural network software.