基于鲁棒性集成模板和动态响应模式的集装箱码头资源分配建模与优化研究

This project proposes to study three interrelated resource allocation problems arising from the operational management of container terminals. After analyzing the complicating characteristics including the co-existence of vessel and container tasks, the integration of seaside and yard resources, and the periodic operational environment, a conceptual model of key resource allocation at gate terminals is established. Based on the conceptual model, to cope with typical uncertainties, a proactive resource-buffering strategy is firstly proposed at tactical decision level, along with a mathematical optimization model for the integrated robust template. Then the dynamics of multi-stage decision process at operational decision level is analyzed, and a dynamic responsive mode is constructed, as well as it two-stage approximation model at each decision node. Lastly, the coupled tactical-operational decision model is proposed based on iterative interaction and approximate evaluation. At the same time, to solve the above complicated problems, this research develops respectively an iterative algorithm framework based on the approximate mapping relationship, and a meta-heuristic algorithm framework based on the variable levels and decision logic. All the studied modeling and solving methods not only extend current theory of container terminal operational optimization, but also provide core support for the scientific decision-making process of the cyclic flexible resource allocation problem, which is commonly existed in actual production and service systems.

本项目从集装箱码头运营管理背景出发，提出了三个相互关联的资源分配决策问题。分析船舶与集装箱任务并存、海侧与堆场资源集成、周期性运营环境等复杂特征，建立集装箱进出口码头关键资源分配的概念模型。在此基础上针对码头典型不确定性因素，首先提出战术层资源分配的前摄缓冲策略，建立鲁棒性集成模板决策的数学优化模型；随后分析运作层多阶段决策过程的动态特性，建立动态响应模式及各节点的二阶段近似优化模型；最后研究基于迭代式交互与近似评价的战术-运作层面耦合建模方法。同时，对于上述复杂问题的优化求解，提出并研究基于近似映射关系的迭代算法框架以及基于变量层级和决策逻辑的元启发式算法框架。以上研究的建模与算法都是对现有码头运营优化理论的扩展与延伸，同时也为一类在实际生产与服务系统中普遍存在的循环式柔性型资源分配问题的科学决策提供核心支撑。

本项目以集装箱码头运营管理为背景，针对码头运营管理过程中周期性、不确定性及多资源调度等决策因素，从战术层与运作层的角度出发，首先建立鲁棒性集成模板决策、动态响应模式及各节点的二阶段近似决策及多资源分配决策等优化模型，随后对于上述复杂问题提出基于近似映射关系的迭代算法框架以及基于变量层级和决策逻辑的元启发式算法框架，为一类在实际生产与服务系统中普遍存在的循环式柔性型资源分配问题的科学决策提供核心支撑。取得的成果在国内外相关领域重要学术期刊发表SCI/EI论文10篇，学术会议论文1篇，并同时共培养5名硕士研究生，为多家企业输送了优秀人才。