Fast Algorithms for Small Area Electric Load Forecasting

Abstract

Advanced techniquesc for forecasting future distribution system loads require computations that normally consume large amounts of computer time. Two of the more common steps in modern forecasting methods are shown to be solvable using signal processing techniques, resulting in computer processing that is much faster. Details of implementation are given, along with comparison tests of the new approach versus traditional techniques.