No abstract

The aim of this paper is to reduce losses and improve voltage profile of a distribution feeder taking into account probabilistic nature of PV sources and loads, using particle swarm optimization algorithm. The proposed method uses normal distribution and beta distribution for probabilistic modeling of the loads and irradiation respectively. By applying probabilistic load flow, voltage profile and losses for the network are cal culated 24 hours a day and four seasons a year. Based on overall results the optimal size of installed PV resources in each buses on the feeder are obtained and compared to the results obtained from a conven tional method.