• 제목/요약/키워드: hybrid empirical method

검색결과 52건 처리시간 0.016초

경기도 대기질 개선 정책의 온실가스 동시 저감 및 그에 따른 공편익 효과 분석 (Greenhouse Gas Reduction by Air Quality Management Policy in Gyeonggi-do and Its Co-benefit Analysis)

  • 김동영;최민애
    • 한국대기환경학회지
    • /
    • 제33권6호
    • /
    • pp.570-582
    • /
    • 2017
  • In recent years, national and local government's air quality management and climate change adaptation policy has been significantly strengthened. The measures in the two policies may be in a relationship of trade-off or synergy to each other. Greenhouse gases and air pollutants are mostly emitted from the same sources of using considerable amounts of fossil fuels. Co-benefits, in which either measure has a positive effect on the other, may be maximized by reducing the social costs and by consolidating the objectives of the various policies. In this study, the co-benefits were examined by empirically analyzing the effects of air pollutants and greenhouse gas emission reduction, social cost, and cost effectiveness between the two policies. Of the total 80 projects, the next 12 projects generated co-benefits. They are 1) extend restriction area of solid fuel use, 2) expand subsidy of low-$NO_x$ burner, 3) supply hybrid-vehicles, 4) supply electric-vehicles, 5) supply hydrogen fuel cell vehicles, 6) engine retrofit, 7) scrappage of old car, 8) low emission zone, 9) transportation demand management, 10) supply land-based electric of ship, 11) switching anthracite to clean fuel in private sector, 12) expand regional combined-energy supply. The benefits of air pollutants and greenhouse gas-related measures were an annual average of KRW 2,705.4 billion. The social benefits of the transportation demand management were the highest at an annual average of KRW 890.7 billion, and followed by scrappage of old cars and expand regional combined-energy supply. When the social benefits and the annual investment budgets are compared, the cost effectiveness ratio is estimated to be about 3.8. Overall, the reduction of air pollutants caused by the air quality management policy of Gyeonggi-do resulted in an annual average of KRW 4,790.2 billion. In the point sources management sector, the added value of $CO_2$ reduction increased by 4.8% to KRW 1,062.8 billion, while the mobile sources management sector increased by 3.6% to KRW 3,414.1 billion. If social benefits from $CO_2$ reduction are added, the annual average will increase by 7.2% to KRW 5,135.4 billion. The urban and energy management sectors have shown that social benefits increase more than twice as much as the benefits of $CO_2$ reduction. This result implies that more intensive promotion of these measures are needed. This study has significance in that it presents the results of the empirical analysis of the co-benefits generated between the similar policies in the air quality management and the climate change policy which are currently being promoted in Gyeonggi-do. This study suggested that the method of analyzing the policy effect among the main policies in the climate atmospheric policy is established and the effectiveness and priority of the major policies can be evaluated through the policy correlation analysis based on the co-benefits. It is expected that it could be a basis for evaluation the efficiency of the climate change adaptation and air quality management policies implemented by the national and local governments in the future.

머신러닝 기반 페로브스카이트 태양전지 광흡수층 박막 최적화를 위한 연구 (A Study on Optimization of Perovskite Solar Cell Light Absorption Layer Thin Film Based on Machine Learning)

  • 하재준;이준혁;오주영;이동근
    • 한국콘텐츠학회논문지
    • /
    • 제22권7호
    • /
    • pp.55-62
    • /
    • 2022
  • 페로브스카이트 태양전지는 4차 산업혁명으로 사물인터넷, 가상환경 등의 증가에 따른 전력 수요가 급증하면서 점진적으로 고갈되어가는 석유, 석탄, 천연가스 등의 화석연료를 대체할 태양에너지, 풍력, 수력, 해양에너지, 바이오에너지, 수소에너지 등의 신재생 에너지 분야에서 연구가 활발한 부분이다. 페로브스카이트 태양전지는 페로브스카이트 구조를 가진 유-무기 하이브리드 물질을 사용하는 태양전지 소자로 고효율, 저가의 용액 및 저온 공정으로 기존의 실리콘 태양전지를 대체할 수 있는 장점들이 있다. 기존의 경험적 방법으로 예측한 광흡수층 박막을 최적화하기 위해서 소자 특성 평가를 통해 신뢰도를 검증해야 한다. 그러나 광흡수층 박막 소자 특성 평가 비용이 많이 소요되므로 시험 횟수에 제약이 따른다. 이러한 문제점을 해결하기 위하여 광흡수층 박막 최적화의 보조 수단으로 머신러닝이나 인공지능 모델을 이용하여 명확하고 타당한 모델의 개발과 적용 가능성이 무한하다고 본다. 이 연구에서는 페로브스카이트 태양전지의 광 흡수층 박막 최적화를 추정하기 위하여 서포트 벡터 머신의 선형 커널, 가우시안 커널, 비선형 다항식 커널, 시그모이드 커널의 회귀분석 모델을 비교하여 커널 함수별 정확도 차이를 검증하였다.