http://chineseinput.net/에서 pinyin(병음)방식으로 중국어를 변환할 수 있습니다.
변환된 중국어를 복사하여 사용하시면 됩니다.
실시간 열량계 정보를 활용한 단기 열 수요 예측 모델 개발에 관한 연구
송상화 ( Sang Hwa Song ),신광섭 ( Kwangsup Shin ),이재훈 ( Jaehun Lee ),정윤재 ( Yunjae Jung ),이재승 ( Jaeseung Lee ),윤석만 ( Seokmann Yoon ) (사)한국빅데이터학회 2020 한국빅데이터학회 학회지 Vol.5 No.2
지역난방 시스템은 서비스 지역 내 열 수요처들을 네트워크로 연결하여 중앙의 저비용 고효율 열 생산설비를 통해 열을 공급하는 에너지 시스템이다. 효율적인 열 공급 시시스템 운영을 위하여 지역 내 열 수요를 정확하게 예측하고 이를 바탕으로 열 생산 계획을 최적화하는 것이 중요하다. 본 연구에서는 지역 내 열수요처별 열 사용량 패턴에 대한 빅데이터 정보로 기계실별 실시간 열량계 정보를 반영한 열수요 예측모형을 제시하였다. 기존에도 열 수요예측에 활용되던 지역 전체 열수요 실적 합계와 함께 수요처별로 설치되어 있는 열량계로부터 실시간으로 수집한 개별 열수요 실적을 예측모형에 반영함으로써 열 수요처별로 상이한 열사용 패턴을 반영한 열 수요 예측이 가능할 것으로 기대된다. 지역난방 기업의 실제 열수요 실적을 바탕으로 열수요 예측 정확도를 측정한 결과 계절에 상관없이 기본 모형 대비 열량계 빅데이터를 반영할 경우 정확도가 올라가는 것으로 분석되었으며, 향후 열수요처별 다양한 형태의 데이터를 추가로 반영함으로써 열 수요 예측 정확도 향상이 가능할 것으로 예츢된다. District heating system supplies heat from low-cost high-efficiency heat production facilities to heat demand areas through a heat pipe network. For efficient heat supply system operation, it is important to accurately predict the heat demand within the region and optimize the heat production plan accordingly. In this study, a heat demand forecasting model is proposed considering real-time calorimeter information from local heat demands. Previous models considered ambient temperature and heat demand history data to predict future heat demands. To improve forecast accuracy, the proposed heat demand forecast model added big data from real-time calorimeters installed in the heat demands within the target region. By employing calorimeter information directly in the model, it is expected that the proposed forecast model is to reflect heat use pattern of each demand. Computational experiemtns based on the actual heat demand data shows that the forecast accuracy of the proposed model improved when the calorimeter big data is reflected.
집단에너지 네트워크의 최적 열 생산 및 연계 시스템 구축에 관한 연구
송상화(Sang Hwa Song),신광섭(Kwang Sup Shin),이재승(Jae Seung Lee),임신영(Shin-Young Im) 대한설비공학회 2016 대한설비공학회 학술발표대회논문집 Vol.2016 No.6
In a district heating and cooling system, various types of production facilities are used to produce hot steam water. These facilities include peak load boilers, incinerators and combined heat and power platns to name a few. Each production facility has their own operation characteristics. Peak load boilers are expensive but can control the production level easily, which is thus useful to deal with unexpected high heat demands. Incinerators are cheaper but the production capacities are comparatively small. They typically produce hot steam water consistently without changing production level dynamically. Combined heat and power plants produce both heat and electricity, while their economic efficiency is mostly dependent on electricity market conditions. In addition, the district heating and cooling system considered in this research is quite complicate since the system consists of multiple nodes, each of which has different portfolios of production facilities. Nodes are interconnected so that hot steam waters are supplied to neighborhood demand nodes when necessary. The main objective of the decision support system developed in this research is to match heat demands with a portfolio of production facilities while maximizing the overall profit of the system.
비선형 연료 제약 및 유지보수 비용을 고려한 Mixed Integer Linear Programming 기반 발전기 주간 운용계획 최적화
송상화(Sang Hwa Song),이경식(Kyung Sik Lee) 한국경영과학회 2008 經營 科學 Vol.25 No.1
This paper considers a profit-based unit commitment problem with fuel consumption constraint and maintenance cost, which is one of the key decision problems in electricity industry. The nature of non-linearity inherent in the constraints and objective functions makes the problem intractable which have led many researches to focus on Lagrangian based heuristics. To solve the problem more effectively, we propose mixed integer programming based solution algorithm linearizing the complex non-linear constraints and objectives functions. The computational experiments using the real-world operation data taken from a domestic electricity power generator show that the proposed algorithm solves the given problem effectively.
Deep Learning 기반 네트워크 열수요 예측에 관한 연구
송상화(Sang Hwa Song),이재승(Jae Seung Lee) 대한설비공학회 2018 대한설비공학회 학술발표대회논문집 Vol.2018 No.6
Heat demand forecasting is to predict short-term or long-term heat demands within a predetermined heat demand area. In order to efficiently operate the district heating system, it is necessary to establish an optimal heat supply plan to predict and supply heat demand precisely and the heat demand forecasting has been actively studied. The previous study on the prediction of heat demand has been based on the characteristics of outdoor air temperature, weather, and heat demand history with regression analysis and neural network approaches. However, As deep learning techniques are widely applied to the data analysis field, this study examines the applicability of deep learning model to the heat demand forecasting. In this study, we examined the deep learning model for short-term heat demand forecasting. The deep learning based heat demand forecasting with the actual heat demand history are analyzed and the results show that the model gives a better prediction model in some heat demand areas than the conventional regression and neural network model. If sufficient data are secured in the future, it is expected that accuracy will be improved.
송상화(Sang Hwa Song),신광섭(Kwang Sup Shin),김재곤(Jae Gon Kim),정진용(Jin Yong Jeong) 한국산업경영시스템학회 2015 한국산업경영시스템학회지 Vol.38 No.1
Korea ocean research stations manage the weather and environmental data collected from coastal and ocean areas to provide short-term and long-term ocean forecasts. The purpose of this paper is to analyze and quantify economic benefits of the ocean research stations with sensors to observe physical, chemical, and biological data. The construction and operation of an integrated ocean observation station is expected to reduce uncertainty about ocean and coastal areas and to improve the quality of ocean forecasts. The economic benefits are mainly come from improved search and rescue operations, ocean pollution management, yellow dust management, and improved productivity in ocean-related industries. In addition, an input-output analysis is performed to evaluate the economic impacts of ocean research stations nationwide. The analysis shows that the system can contribute to industries such as fishing, maritime and air cargo, medical and health care.