不确定与动态信息环境下基于预规划-重规划集成建模的应急物流选址-调度鲁棒优化研究

项目名称： 不确定与动态信息环境下基于预规划-重规划集成建模的应急物流选址-调度鲁棒优化研究

项目编号： No.71471007

项目类型： 面上项目

立项/批准年度： 2015

项目学科： 管理科学

项目作者： 周泓

作者单位： 北京航空航天大学

项目金额： 61万元

中文摘要： 紧急救援是减轻灾害性突发事件后果的首要任务之一，而科学的应急物流管理则是实现高效救援的基础。由于实际环境中存在大量的不确定因素和各种动态变化，使得原先安排好的方案可能无法顺利实施，有效解决手段是寻找一个高鲁棒性的预规划，并在实际运作过程中针对发生的变化和新信息对初始规划进行修正（重规划）。预规划-重规划是相互影响的，只有实现二者的协同优化才能达到最佳效果。本项目针对应急物流设施选址和车辆调度的集成鲁棒优化问题拟在两个层面实现协同优化，即预规划-重规划（包括选址-再选址与调度-重调度）的集成，以及设施选址与车辆调度的集成。总体思路是采用递阶优化架构，以鲁棒优化实现上层预规划，以反应式策略修复实现下层重规划；上层以鲁棒优化结合智能启发式搜索算法为主要手段，下层以各种启发式修复规则为主要手段，利用系统仿真过程将二者加以衔接，从而建立一套应急物流管理的新型、高效、实用鲁棒优化方法。

中文关键词： 物流管理；应急管理；鲁棒优化；集成优化；选址-调度

英文摘要： Emergency rescue is one of the top tasks to relieve the effects of disastrous events, while proper emergency logistics management plays an essential role it the process. Due to the enormous uncertainty and different dynamic changes in the environment, it is quite usual that the initial planning becomes invalid finally. An effective way to solve this problem is to find a highly robust pre-planning and to make the reactive revision against the changes and new information (re-planning). There is no doubt that pre-planning and re-planning have strong interaction between each other, and the best result can be reached only when there is a proper coordination between them. The robust optimization for location/scheduling of emergency logistics management based on pre-planning/re-planning integrated modelling with uncertain and dynamic information is targeted in this project, with the aim of implementing two kinds of effective integrations, the first is. pre-planning and re-planning (including the integration of location and relocation, as well as scheduling and rescheduling), the second is location and scheduling. The main idea is to adopt a hierarchical modelling framework, in which the upper pre-planning is implemented with a robust optimization procedure while the lower re-planning is handled by reactive repairing for the previous strategies. The sophisticated meta-heuristic methods will be employed and incorporated with robust optimization procedure, and some priority-based heuristic rules will be developed for strategy repairing in order to ensure the reasonable computation efficiency for the optimization. Simulation will be employed to link the two levels. A set of novel, effective, and practical robust optimization methods for emergency logistics management is expected to be developed at the end of the project.

英文关键词： logistics management;emergency management;robust optimization;integrated optimization;location-scheduling