| Title |
Evaluation of Weather Information in Forecasting Daily Peak Load of Electricity Demand |
| Authors |
Ju-Yeol Ryu ; Jae-Min Cha ; Bo-Ra Lee |
| DOI |
http://dx.doi.org/10.5207/JIEIE.2018.32.12.073 |
| Keywords |
Daily Electricity Demand Peak Load ; Forecasting ; Weather Information ; Heating And Cooling Degree Day ; Holiday Effect |
| Abstract |
Daily electricity demand forecasting for peak load is an important indicator for domestic electricity power plant operation. In this paper, we tried to improve the predictive accuracy by using weather information which is widely used for electricity demand forecasting. We analyzed the effect of daily-based temperature, wind speed, and humidity and especially compared the effect of temperature with the national representative temperature, which was calculated by weighted mean of the five major cities in Korea, Cooling Degree Day (CDD), Heating Degree Day (HDD), and Heating and Cooling Degree Day (HCDD), which was derived from the combination of HDD and CDD. After fitting the training data from 2010 to 2016 and testing the several models for the year 2017 in order to evaluate the predictive accuracy of each model, the model including HCDD and holiday effect showed the best performance based on Mean Absolute Percentage Error. |