Isolated area load forecasting using linear regression analysis: practical approach

This paper presents an analysis to forecast the loads of an isolated area where the history of load is not available or the history may not represent the realistic demand of electricity. The analysis is done through linear regression and based on the identification of factors on which electrical load growth depends. To determine the identification factors, areas are selected whose histories of load growth rate known and the load growth deciding factors are similar to those of the isolated area. The proposed analysis is applied to an isolated area of Bangladesh, called Swandip where a past history of electrical load demand is not available and also there is no possibility of connecting the area with the main land grid system.