Short-Term Load Forecasting Approach Based on Different Input Methods of One Variable: Conceptual and Validation Study

Electrical demand forecasting is a key element within the electrical power system. STLF is considered the most significant for many processes in the Power Grid. A slight improve in the STLF accuracy may fetch a lot of benefits at both levels; enhances the grid stability, and increases economic benefits. The aim of the study is to find a simple accurate short term forecasting method. The proposed method based on the proposed concept that recognizes the many relevant factors affect the load in the short term. Lack of information about the relations between load and those factors have led to unclear methodology to deal with this uncertainty. Thus, STLF can rely on Historical Load Data only, as an observer for all influence factors. The proposed approach is implemented, to be verified, via Artificial Neural Network (ANN), by Matlab software, on four different inputs methods, all use one variable. The obtained result is auspicious. It reduces the Mean Absolute percentage Error dramatically compared to the previous methods.