제목 강건한 선석 계획을 위한 가우시안 혼합 모델 기반의완충 시간 적용
Title Robust berth scheduling with GMM-based buffer time
저자 우성훈 (현대모비스 데이터사이언스팀)
박현지 (한국해양과학기술원 부설 선박해양플랜트연구소 해양안전환경연구본부)
전성우 (고려대학교 산업경영공학부)
김아름 (한양대학교 교통물류공학과)
조성원* (한국해양과학기술원 부설 선박해양플랜트연구소 해양안전환경연구본부)
Author Sung Hun Woo(Data Science Team, Hyundai Mobis)
Hyun Ji Park(Maritime Safety and Environmental Research Division, Korea Research Institute of Ships and Ocean Engineering)
Sungwoo Jun(School of Industrial and Management Engineering, Korea University)
Armi Kim(Department of Transportation and Logistics Engineering, Hanyang University)
Sung Won Cho†(Maritime Safety and Environmental Research Division, Korea Research Institute of Ships and Ocean Engineering)
Bibliography Journal of Logistics Science & Technology, 3(1),1~21, 2022,
DOI
Key Words Container terminal, Berth scheduling, GMM(Gaussian mixture model), Buffer time
Abstract With global changes as the spread of COVID-19, resource management of container terminals has become more difficult. For this reason, terminal managers have to come up with robust schedules to manage vessels, and many research have conducted to provide robust berth schedule fromdisruptions such as vessel delays. In this paper, a two-phase mathematical model is proposed to generate a robust berth schedule with GMM(gaussian mixture model)-based buffer time. In phase 1, tardiness cost is minimized and the best berthing position is determined. In phase 2, the schedule is rescheduled to minimize the deviation from the baseline schedule obtained by phase 1. At the same time, in order to assign optimal buffer time to each vessel, historical data is analyzed and applied to the proposed model by utilizing machine learning techniques. The proposed method, compared to conventional methods, showed improvement in reducing the total cost incurred for a terminal when encountered with disruptions.
PDF download JM_3[1]-P1~21.pdf