한국해양대학교

Detailed Information

Metadata Downloads

Port Productivity Analysis based on Automation Level using Deep Learning Method

DC Field Value Language
dc.contributor.advisor 김환성 -
dc.contributor.author QU ZITONG -
dc.date.accessioned 2024-01-03T17:28:42Z -
dc.date.available 2024-01-03T17:28:43Z -
dc.date.created 2023-03-03 -
dc.date.issued 2023 -
dc.identifier.uri http://repository.kmou.ac.kr/handle/2014.oak/13140 -
dc.identifier.uri http://kmou.dcollection.net/common/orgView/200000671295 -
dc.description In recent years, due to the increasing influence of global economic integration, international trade and economy have made indelible contributions to the economic development of the region and the whole country. As an important node connecting sea and land transportation, the development of import and export trade has an increasing requirement on port supply capacity and port service efficiency, which also promotes the continuous expansion of the scale of ports. With the implementation of the strategy of maritime power in recent years, the shipping industry has been revitalized with the development of the new Maritime Silk Road. Throughput is an important technical index of container logistics, no matter which port has gradually tried to manage and predict its port throughput in a more scientific way. Up to now, container logistics has almost occupied a dominant position in the whole shipping logistics, and the research on throughput is more and more important. As one of the important ports in the North of China, Qingdao Port has diversified advantaged advantages, and has become an important town in the container shipping system in the North of China. Since 2008, some major hubs at home and abroad began to put forward the construction concept of "port automation", and the research on "port automation" has also become a research hotspot. In the research on "port automation" as the theme, most of the research focuses on the connotation and development direction of port automation and the application and prospect of automation technology in the process of port operation. In view of this, this paper tries to summarize the existing research results, and from the perspective of automation technology, analyze the automation development status of Qingdao Qianwan Container Terminal and the influencing factors of port throughput. LSTM algorithm was used to build the port throughput prediction model, and different port automation index was used as the model input to carry out the target port throughput prediction experiment. The model with high throughput prediction accuracy was obtained to systematically analyze and experiment with the throughput trend of Qingdao Qianwan Container Terminal. The port throughput prediction model based on LSTM is constructed, and the target port throughput prediction experiment is carried out through different port automation indexes, which proves the applicability and prediction accuracy of the LSTM algorithm model in port throughput prediction. At the same time, according to the forecast results, the port development trend analysis, provides a basis for the planning and construction of Qingdao Qianwan Container Terminal. -
dc.description.abstract 최근 몇 년간 세계 경제 통합의 영향력이 증가함에 따라, 국제 무역과 경제는 지역과 국가 전체의 경제 발전에 지울 수 없는 기여를 해왔다. 해상과 육상 교통을 연결하는 중요한 노드로서 수출입 무역의 발전은 항만 공급 능력과 항만 서비스 효율성에 대한 요구가 증가하고 있으며, 이는 항만의 지속적인 규모 확대를 촉진하기도 한다. 항만에서 항만 처리량을 보다 과학적인 방법으로 관리하고 예측하는 것은 컨테이너 물류의 중요한 기술 지표로 되고 있다. 지금까지 컨테이너 물류는 해운물류 전체에서 거의 지배적인 위치를 차지하고 있으며, 처리량에 대한 연구가 점점 더 중요해지고 있다. 칭다오 항은 중국 북부의 중요한 항구 중 하나로서, 컨테이너 물동량의 다양화로서 중국 북부의 컨테이너 운송측면에서 중요한 도시가 되었다. 칭타오항은 2008년부터 국내외 일부 주요 거점들이 '항만자동화'라는 구축 개념을 내세우기 시작했고, '항만자동화'에 대한 연구가 주요 이슈로 자리 잡았다. '항만자동화' 주제의 연구에서는 항만자동화의 발전방향, 항만운영과정에서의 자동화 기술의 적용과 전망에 대한 연구가 대부분이다. 이에, 본 논문은 기존의 연구 결과를 요약하고, 자동화 기술의 관점에서 칭다오 첸완 컨테이너 터미널의 자동화 개발 현황과 항만 처리량의 영향 요인을 분석하고자 한다. LSTM 알고리즘을 사용하여 항만 물동량 예측 모델을 구축하고, 모델 입력으로는 서로 다른 항만 자동화 지수를 사용하여 대상 항만 물동량 예측 시뮬레이션을 수행하였다. 칭다오 첸완 컨테이너 터미널의 처리량 추이를 체계적으로 분석하고 실험을 통하여 물동량 예측 정확도가 높은 모델을 얻었다. 동시에, 예측 결과에 의한 항만 개발 동향 분석으로 칭다오 첸완 컨테이너 터미널의 계획과 건설에 대한 근거를 타당성을 고찰할 수 있었다. -
dc.description.tableofcontents Chapter 1 Introduction 1 1.1 Background 1 1.2 Motivation for research 4 1.3 Contributions 5 1.4 Organization of this dissertation 8 Chapter 2 Previous research 9 2.1 Overview of Port automation development 9 2.2 Qingdao Port Company profile 13 2.3 Automatic construction of Qingdao Qianwan Container Terminal 14 Chapter 3 Methods 17 3.1 Long Short-Term Memory 17 3.1.1 Deep Learning 17 3.1.2 Recurrent Neural Network 18 3.1.3 Research review on LSTM prediction application 18 3.1.4 Principle of LSTM Neural network 21 3.1.5 Bi-directional Long Short-Term Memory 25 3.2 Port Automation Index 26 3.2.1 Yard 28 3.2.2 Quay 32 3.2.3 Transfer 37 3.2.4 Gate 40 Chapter 4 Container throughput forecast of Qingdao Qianwan Container Terminal combined with port automation level 45 4.1 Qingdao Qianwan Container Terminal port automation level 45 4.1.1 Qingdao Qianwan Container Terminal Yard automation level 45 4.1.2 Qingdao Qianwan Container Terminal Quay automation level 47 4.1.3 Qingdao Qianwan Container Terminal Transfer automation level 49 4.1.4 Qingdao Qianwan Container Terminal Gate automation level 50 4.1.5 Qingdao Qianwan Container Terminal port automation level 51 4.2 Prediction model construction of Qingdao Qianwan Container Terminal 52 4.2.1 Model building 52 4.2.2 LSTM test results 53 4.2.3 Forecasting 55 Chapter 5 Discussion 60 Reference 62 -
dc.format.extent 69 -
dc.language eng -
dc.publisher 한국해양대하교 대학원 -
dc.rights 한국해양대학교 논문은 저작권에 의해 보호받습니다. -
dc.title Port Productivity Analysis based on Automation Level using Deep Learning Method -
dc.title.alternative 항만 자동화 수준 기반의 항만 물동량 딥러닝 분석 -
dc.type Dissertation -
dc.date.awarded 2023-02 -
dc.embargo.terms 2023-03-03 -
dc.contributor.alternativeName 곡자동 -
dc.contributor.department 대학원 물류시스템학과 -
dc.contributor.affiliation 한국해양대학교 대학원 물류시스템학과 -
dc.description.degree Master -
dc.identifier.bibliographicCitation QU ZITONG. (2023). Port Productivity Analysis based on Automation Level using Deep Learning Method. -
dc.subject.keyword Deep learning, Long Short-Term Memory, Qingdao port, Port Automation, Port Automation Index -
dc.identifier.holdings 000000001979▲200000003272▲200000671295▲ -
Appears in Collections:
기타 > 기타
Files in This Item:
There are no files associated with this item.

Items in Repository are protected by copyright, with all rights reserved, unless otherwise indicated.

Browse