FORECASTING ELECTRICITY CONSUMPTION USING REGRESSION ANALYSIS

This research focused on forecasting monthly average electricity consumption in Rayong province, Thailand, using regression analysis method with 3 factors, monthly average number of electricity customers, monthly average rainfall and monthly average temperature. Data of 80 samples was divided into two groups; the first group was to build a regression model and the other was to validate the performance of forecasting electricity consumption. After fitting the model, assumptions of regression analysis were detected by Anderson-Darling statistic, Durbin-Watson statistic, Breusch-Pagan statistic and Variance inflation factor. Finally, the performance of forecasting electricity consumption values was calculated by mean absolute percentage error with 0.0416, mean absolute error with 407,061.084 and root mean squared error with 596,706.8834. The results illustrated that the regression standard error was 231,204 with the adjusted coefficients of determination of 0.931.