| Summary: | Every year, the demand of electricity is always increased. It is due to the effect of population increasing. Its demand requires the electricity providers could distribute the electricity needs of consumers so that their activities do not have problems. In fact, the electricity is generated by power plants can not be saved, but it is used directly by consumers. Because of ever-changing power needs, it is required load forecasts or forecasts of consumer power needs as a basic for planning operations. The purpose of this study is to examine whether the double seasonal exponential smoothing method is suitable for daily electricity load forecasting in Central Java and Yogyakarta. Double Seasonal Exponential Smoothing method (DSES) is capable of calculating seasonal and trends that occur in electricity load data. The selection of smoothing constant value is very important in forecasting using DSES. With the help menu add-ins \"Solver\" in Microsoft Excel smoothing constant values obtained are capable of producing a minimum value of forecasting error.
Based on research results represent that the forecasting results using the DSES method has a high accuracy rate. Even its method is able to generate load forecasting DSES with higher accuracy than using the coefficients are carried out by PLN. Load forecasting using DSES method have value of MAPE 1.588392% and MSE 2221.6866. While the coefficient method have value of MAPE 4.565433% and MSE 15018.64. However, load forecasting method using DSES are only suitable used for forecasting the load of work and are not recommended for predicting the load on weekends.
|
|---|