摘 要

当前，我国地铁建设正处于大规模高速发展时期。然而，地铁施工事故频发，造成了巨大的财产损失和人员伤亡，给地铁工程的应急管理工作带来了极大的挑战。应急物资调度决策作为应急管理的关键环节，对于控制事态发展、减少灾害损失、提高救援效率具有重要的意义。因此，本文研究地铁施工事故背景下的应急物资动态调度决策模型具有重要的理论和实践意义。

本文首先介绍了地铁施工事故和应急物资调度的相关理论知识。其次，在考虑了动态行程时间和动态需求两个动态要素的基础上，以应急响应时间最早和出救点数目最少为优化目标，构建了实时动态的多阶段应急物资调度模型，该模型能有效解决事故发生后出救点组合优化问题。针对动态行程时间和动态需求，分别提出了基于浮动车理论的行程时间估计方法和滚动时域策略进行求解，保障了其作为动态调度模型输入条件的实时性，确保了模型运行结果的科学有效。然后，针对提出的应急物资动态调度模型，通过滚动时域策略将动态调度问题转化为一系列离散时刻点的静态调度问题，并设计了改进遗传算法进行求解。在理论分析的基础上，以某地铁施工事故为案例，通过MATLAB编程对算法进行了实现，同时通过与标准遗传算法运算结果比对分析，更加验证了模型的可行性和算法的科学高效性。最后，以理论研究成果为核心，设计和开发了地铁施工应急物资调度系统，通过系统功能展示，验证了系统的可靠性。

本研究涉及管理科学与工程、交通运输学、运筹学、计算机等多个学科知识，提出了解决地铁施工应急物资动态调度问题的决策模型和求解策略，丰富了应急管理理论。同时，开发了应急物资调度系统，为地铁施工应急救援提供辅助决策支持，提高了应急管理能力，具有重要的实用价值。

关键词：地铁施工事故应急物资动态调度模型动态行程时间改进遗传算法

Abstract

At present, the metro construction of China is in a large-scale rapid development period. However, the frequent accidents results in a huge loss of property and casualties, which has brought great challenges to the metro project emergency management. As the key link of emergency management, emergency material scheduling decision-making makes great significance to taking situations under control, reducing the loss of disaster and improving the efficiency of rescue. Therefore, it is of great theoretical and practical significance to study the dynamic scheduling decision model of emergency material scheduling.

This research firstly introduces the related theoretical knowledge of the metro construction accidents and the emergency material scheduling. Secondly, based on the two dynamic elements of travel time and demand, the real-time dynamic multi-stage emergency material scheduling model is constructed with the earliest emergency response time and the number of rescue points.The model can effectively solve the problem for combinatorial optimization of multi-depot selection after the event of an accident. The travel time estimation method based on the floating vehicle theory and the rolling horizon strategy are proposed to solve the dynamic travel time and demand, which ensures the punctuality of dynamic factors as a scheduling model input conditions and makes model results more scientific and effective. Then, the dynamic scheduling problem is transformed into a series of discrete time points by rolling horizon strategy. And an improved genetic algorithm is designed to solve the dynamic scheduling model of emergency materials. On the basis of the theoretical analysis, the algorithm is implemented by MATLAB programming. A metro construction is taken as an example to verify the feasibility of the model and the algorithm. At the same time, compared with the standard genetic algorithm, improved genetic algorithm has greater superiority in computing efficiency and better operation results. Finally, with the theoretical research results as the core, this thesis designed and developed the metro construction emergency material scheduling system. Through the system function display, the system reliability is verified.

This study involves a number of disciplines such as management science and engineering, transportation, operations research, computer and so on, and puts forward the decision model and solution strategy to solve the dynamic scheduling problem of metro construction emergency events and enrich the emergency management theory. At the same time, the emergency material scheduling system is developed to provide supportive decision support for the metro construction emergency rescue, which contains important practical value to improve the emergency management capabilities.

Keywords:metro construction accidentemergency materials dynamic scheduling model dynamic travel-time improved genetic algorithm