Automated Short-Term Load Forecasting Using Modified Stochastic Hour Ahead Proportion (SHAP) Analysis

The accuracy of the electricity demand forecast is important in the operations of a power system. The objective of this paper is to generate an automated forecast model with better accuracy than its predecessors. The Automated ShortTerm Load Forecasting Using Modified Stochastic Hour Ahead Proportion (SHAP) Analysis is generated using the C# program. The constructed application would then use an excel spreadsheet to input data. Therefore, the data would be used as a basis for the construction of the forecast model. The output would then display the equation and the graph of the constructed forecast model. Based on the output of the program, the generated forecast model has a MAPE and SDE value of 0.941725 and 1.149855 respectively. The MAPE and SDE value of the generated forecast model is significantly lower than the MAPE and SDE values of the SHAP, WMA, SHAP-WMA which are 3.765681 and 2.822254, 5.610123 and 10.312887, 3.278946 and 3.055406 respectively. This comparison shows that the Modified SHAP Analysis is statistically better than its predecessors.