不确定集港环境下班轮航线的动态配载决策方法研究

The stowage decision-making in liner route is closely relative with maritime shipping profit of containerized shipping supply chain. So far, exiting stowage decision theories lack the research on efficiency degeneration of stowage in liner route. Stowage plans produced by exiting stowage decision theories have large frangibility in uncertain environment. Our project is aim to research and solve the dynamic stowage decision problem in liner route with uncertain collection process of export containers, which is characterized by 4-dimensions container loading, no aftereffect and efficiency degeneration of stowage along liner route. Firstly, we explore the multi-stages decision modeling and dimensionality reduction approaches of stowage decision in liner route based on 4-dimensions container loading and no aftereffect, and set up multi-stages stowage decision model. Then, analyze how probability events from uncertain environment affect stowage decision in liner route during the uncertain collection process of export containers along liner route, and develop the dynamic stowage decision model for improving the robust of multi-stages stowage decision model. In order to effectively solve the scientific problem presented by our project and to avoid efficiency degeneration of stowage in liner route, we design the adaptive optimizing frame integrating pre-decision of dynamic stowage in liner route and the repairing with itself, and correspondingly develop an adaptive decision algorithm based on the adaptive optimizing frame above. Finally, we will confirm these theories and methods presented by us with aids of comparison experiment, computational simulation and empirical study. Our research project can systematically give dynamic stowage decision approaches in liner route with uncertain collection process of export containers along liner route. It not only enriches uncertain decision methods of stowage in liner route, but also satisfies the need of container maritime effective shipping on dynamic stowage decision in liner route.

在集装箱航运供应链中，班轮配载与集装箱供应链的航运利益密切相关。已有的配载决策理论缺乏班轮与航线之间配载效率退化的研究，配载决策计划易受到不确定概率事件影响，隐含极强脆性。本项目旨在研究并解决不确定集港环境下带有四维装箱、无后效性和配载效率退化的班轮航线动态配载决策问题。针对四维装箱和无后效性等特点，研究班轮航线配载的多阶段决策降维建模方法；分析不确定集港环境的概率事件对班轮航线配载决策的影响，研究班轮航线的动态配载决策建模方法，实现班轮航线配载的鲁棒吸收；为有效求解问题，研究班轮航线动态配载预决策及其修复的自适应集成优化机制，设计能有效收敛的自适应决策算法，避免配载效率退化；运用计算机仿真、比对与实证相结合的实验手段，验证项目所提出的理论与方法。本研究可形成不确定集港环境下班轮航线的动态配载决策方法，丰富班轮航线配载的不确定性决策理论，满足集装箱海洋高效运输对班轮航线动态配载的需要。

国内外贸易离不开集装箱运输，集装箱物流供应链在国民经济中发挥着重要作用，其高效、安全是国家战略“交通强国”中的重要组成部分。集装箱物流供应链中，港口、船舶、航线、资源等要素的有效组织与优化协调，对于充分发挥集装箱供应链的效能至关重要。.集装箱班轮配载与集装箱供应链的航运利益密切相关。集装箱班轮按照特定航线访问各个挂靠港，实现集装箱流在港口间的流动，完成货物在空间上的移动和港口节点上的装卸堆存。已有的配载决策理论缺乏班轮与航线之间配载效率退化的研究，配载决策计划易受到不确定概率事件影响，隐含极强脆性。.本项目旨在研究并解决不确定集港环境下带有四维装箱、无后效性和配载效率退化的班轮航线动态配载决策问题。针对四维装箱和无后效性等特点，研究班轮航线配载的多阶段决策降维建模方法，构建了多阶段决策模型；分析不确定集港环境的概率事件对班轮航线配载决策的影响，研究班轮航线的动态配载决策建模方法，构建了不确定集港环境下班轮航线配载模型，实现班轮航线配载的鲁棒吸收；为有效求解问题，研究班轮航线动态配载预决策及其修复的自适应集成优化机制，设计能有效收敛的自适应决策算法，运用计算机仿真、比对与实证相结合的实验手段，验证了所提出的理论与方法具有较好的鲁棒性。.本项目在国家自然科学基金委的支持下发表和录用论文25篇，SCI、EI收录13篇；获得发明专利1项、实用新型专利1项、软件著作权3项，受理3项发明专利，获湖北省科技进步2等奖1项；培养研究生10名，参加国内外学术交流11次。项目研究可形成不确定集港环境下班轮航线的动态配载决策方法，丰富班轮航线配载的不确定性决策理论，满足集装箱水上高效运输对班轮航线动态配载的需要，为智慧港口的船舶智慧配载和生产组织优化等提供新的思路和可能的解决方案。