Electricity Demand Forecasting for Champasak Province in Lao PDR Using Winter’s Method with Optimizing Level, Trend and Seasonality Smoothing Constant

Abstract

This research aimed to forecast the electricity consumption demand for Champasak province, Lao PDR. The forecasting process started by collecting electricity consumption in the past which were time-series data. Then, the autocorrelation coefficient was calculated and plotted to investigate the pattern of data. The graph showed that trend and seasonal were integrated in the pattern therefore Winter’s method was selected to be an appropriate forecasting model in this research. All parameters in Winter’s method were optimized by using solver function in spreadsheet software. The result of optimization revealed that α = 0.8711, β= 0.1812 and γ = 0.7500 with minimum MAPE = 5.42%. Finally, improved Winter’s forecasting model with optimized parameters was used to forecast the electricity demand for Champasak province, Lao PDR in 2016. The final result showed that the MAPE of proposed model was 5.26% which less than current forecasting method. This bring about electricity planning will be more accurate.