Forecasting uncertainty of Thailand's electricity consumption compare with using artificial neural network and multiple linear regression methods

In this paper, the accurate electricity consumption forecasting has become important decisions in the energy planning of the developing countries. Last decade has several new techniques are used for electricity consumption forecasting to accurately predict the future demand. The considerable amount of electricity consumption modeling was efforts. This research approach to develop electricity models, statistical approach is a good to engineering approaches when observed and measured data is available. The statistical models, linear regression analysis has shown promising results because of the reasonable accuracy and relatively simple implementation which compared to other methods. In this study, artificial neural network and multiple linear regression analysis were performed data from Electricity Generating Authority of Thailand. In the models, gross electricity generation, installed capacity, gross domestic products (GDP) and population are used as independent variables using historical data from 1993 to 2015. Forecasting results are compared using MAPE and RMSE for the test period data. The results indicate electricity consumption model are accurate and minimum cost for electricity generation in Thailand.