Browse > Article
http://dx.doi.org/10.15266/KEREA.2020.29.4.419

Daily Gas Demand Forecast Using Functional Principal Component Analysis  

Choi, Yongok (School of Economics, Chung-Ang University)
Park, Haeseong (Department of Economics, Sungkyunkwan University)
Publication Information
Environmental and Resource Economics Review / v.29, no.4, 2020 , pp. 419-442 More about this Journal
Abstract
The majority of the natural gas demand in South Korea is mainly determined by the heating demand. Accordingly, there is a distinct seasonality in which the gas demand increases in winter and decreases in summer. Moreover, the degree of sensitiveness to temperature on gas demand has changed over time. This study firstly introduces changing temperature response function (TRF) to capture effects of changing seasonality. The temperature effect (TE), estimated by integrating temperature response function with daily temperature density, represents for the amount of gas demand change due to variation of temperature distribution. Also, this study presents an innovative way in forecasting daily temperature density by employing functional principal component analysis based on daily max/min temperature forecasts for the five big cities in Korea. The forecast errors of the temperature density and gas demand are decreased by 50% and 80% respectively if we use the proposed forecasted density rather than the average daily temperature density.
Keywords
Forecasting gas demand; Functional principal component analysis; Changing seasonality; Daily gas demand; Forecasting temperature density;
Citations & Related Records
연도 인용수 순위
  • Reference
1 김인무.김창식.박성근, "에너지 상대가격 변화에 따른 에너지 수요 예측", 경제학연구, 제59권 제1호, 한국경제학회, 2011. 3.
2 김점수.양춘승.박중구, "한국 도시가스용 천연가스의 소비함수에 대한 실증분석- 시간변동계수(TVC) 시계열모형 활용 -", 에너지공학, 제20권 제4호, 한국에너지학회, 2011. 12.
3 박성근, "전력수요의 기온반응 변화 추정과 시사점", KIET 산업경제, 2016. 5.
4 양준모.정민수.조인숙, "도시가스 수요예측 모형에 대한 소고", Journal of the Korean Data Analysis Society,제19권 제6호, 한국자료분석학회, 2017. 12.
5 이근철.한정희, "선별적 샘플링을 이용한 국내 도시가스 일별 수요예측 절차 개발", 한국산학기술학회논문지, 제16권 제10호, 한국산학기술학회, 2015. 10.
6 이성로, "수도권과 지방권 수요예측모형을 통한 전국 도시가스수요전망의 예측력 향상", 자원.환경경제연구, 제 26권 제4호, 한국환경경제학회.한국자원경제학회, 2017. 12.
7 이승재.어승섭.유승훈, "시계열 자료를 이용한 도시가스의 수요함수 추정", 에너지공학 제22권 제4호, 한국에너지학회, 2013. 12.
8 한국전력거래소, "최근 이상기후를 반영한 냉난방부하 분석 및 예측 모형 개발", 외부용역 연구보고서, 2020.
9 Bosq, D.,"Linear Processes in Function Spaces: Theory and Application," Springer-Verlag New York, 2000.
10 Chang, Y., Y. Choi, C. S. Kim, J. I. Miller, and J. Y. Park, "Forecasting Regional Long-Run Energy Demand: A Functional Coefficient Panel Approach," Working Papers 1915, Department of Economics, University of Missouri, 2019.