동적계획법을 이용한 산업부문 에너지사용량 조사 최적화 방안 연구

The KEA (Korea Energy Agency) annually releases "Industry sector energy and GHG emission statistics" by conducting an annual survey for 100,000 facilities. It is a key challenge in the survey to maximize the response rate within a given budget. The KEA has employed face-to-face and web surveys methods and has to decide which one of these two methods would be the best for maximizing the response rate at the end of the survey year. Despite of expensive cost, the face-to-face survey method may possibly have an immediate response from survey participants. On the contrary, the web survey is less expensive but have low response rates. This paper suggests an optimal survey strategy for a large-scale survey on energy consumption and greenhouse gas emissions in the industrial sector. Using a dynamic programming approach, we explore which survey method would be the best at each survey time point. Numerical analysis results show that conducting consecutive rounds of web surveys can achieve a satisfactory response rate within the current budget. However, to achieve a response rate of 90% or more, utilizing the face-to-face survey and sufficient budget allocation are required.