Distribution networks load forecasting using improved clustering method with particular software

This article presents an optimized algorithm for long-term electric load forecast in distribution or transmission network by using appropriate pattern curves. This method is related to the load growth pattern and annual information of planned year on different scenarios of load growth based on non-uniform areas, where high density areas are located in smaller zones, and low-density areas in larger zones. By using appropriate software the most favorable pattern curve for each cell is fitted. These curves are used to forecast the future annual loads. Forecasting of specific cases also has been considered. The resulted data have been applied to comprehension design of a sample city at the end.