【作者】 陈钢铁；

【导师】 帅斌；

【作者基本信息】 西南交通大学 ， 交通运输规划与管理， 2013， 博士

【摘要】 由于危险品具有易燃、易爆、毒害等特殊物理和化学特性,一旦承运这些货物的车辆在运输过程中发生泄漏、火灾或爆炸等事故,不仅会给事故沿线居民的生命健康和财产安全造成不可挽回的伤害,还有可能对周围生态环境造成重大损失。因此,如何确保道路危险品运输安全,降低事故风险,对保障人民群众生命财产安全及社会稳定具有重要理论和实际意义。论文首先对路网的可靠性以及危险品运输路网的可靠性进行分析,接着研究危险品运输网络可靠性失效的形成机理,以及影响危险品运输路网可靠性的五个因素：人的不安全行为、危险品理化特性、机械的不安全状态、环境因素、安全管理因素,最后从预防的角度研究运输路径的先期优化原理。研究危险品运输过程中,政府相关的监管部门与危险品运输企业博弈问题。由于信息不对称,在博弈过程中,政府与危险品承运人经过多次博弈最后达到平衡。对危险物品运输网络的设计进行了分析,政府和危险品运输企业的双层约束条件下,在此基础上兼顾政府期望的风险和危险品运输企业期望的成本上不同的利益目标,建立一个双层规划模型,根据非合作博弈论的相关理论设计启发式算法进行求解。危险品运输路网破坏者和危险品运输承运人可靠性博弈模型,斯塔克伯格-纳什模型和部分合作性博弈模型,对他们的联系和功能进行了分析和验证；给出在什么样的情况下不合作博弈可以给出最坏情况下的解决方案；(1)破坏者进行联合后对危险品运输路网出行成本可靠性的影响(2)如果在路网中加一条路线,所有用户的期望出行成本和出行成本的可靠性都变得更糟的悖论现象；接着在通过从投资方面提高危险品运输路网可靠性方面提出了一个二阶段随机规划,第一阶段找出需要投资的危险品运输路段,第二阶段危险品运输路网中O D对之间的最小遍历成本,可靠性成为了目标函数的因子。利用最短路径的算法把原来的问题推导成另外一种等价形式。通过放松对投资变量的整数限制和对泰勒展开式的应用,得到一个多元线性函数,即为投资边际效益问题。最后对危险品发生事故后从人员疏散、车辆疏散、应急物资调度优化方面建立模型,并通过相关算法对模型进行求解。更多还原

【Abstract】 Dangerous goods are often flammable, explosive, or poisonous. Accidents like leak, catching fire or explosion of dangerous goods in transport may cause huge damage to environment and lives and properties of nearby residents. Therefore, the study on safe transport of dangerous goods has great theoretical and practical significance in protection of people’s lives and property.This thesis analyzes the reliability of transport network, and then explores the mechanism of fail of the transport network of dangerous goods. Furthermore, the five factors influencing reliability of the transport network of dangerous goods are studied, the unsafe behavior of human, physical and chemical properties of dangerous goods, unsafe conditions of machinery, environment, safety management respectively. In the end, the prophase optimization principle of transport paths is studied from the prevention perspective.In the study of dangerous goods transport, the game between the government and dangerous goods carriers is investigated. Owing to asymmetric information in the game process, achieving balance needs several games between the government and the dangerous goods carriers. The thesis designs the transport network of dangerous goods. Considering the different profit targets of the expectation risk of the government and the expectation cost of transport companies, a bi-level programming model is proposed based on the bi-level constraints of the government and the dangerous goods carriers, which is solved by using a heuristic algorithm based on non-cooperative game theory.The thesis analyzes and validates the connections and functions among the reliability game model between destroyers of dangerous goods transport network and carriers of dangerous goods, and the Stackelberg-Nash model, and the part of the cooperative game models. It shows a worst-case solution can be given that in the following non-cooperative games;(1) destroyers influence the reliability of the travel cost of the transport network of dangerous goods after they united (2) add a line in the road network, the reliability of all users’expect travel costs and travel costs have become worse paradox; then to improve the reliability of the transport work of dangerous goods through investment, I proposed a two-stage stochastic programming. The first stage is to find the transport segment of dangerous goods needs to be invested. The second stage is to determine the transport of dangerous goods road network OD pairs between the minimum traversal cost. The reliability factor of the objective function. Use the shortest path algorithm to derive the original problem into another equivalent form. By relaxing the integer restrictions on the investment variable and using the Taylor expansion, a multi-linear function is obtained, that is the marginal effectiveness of investment.In the end, the model is proposed from the perspectives of personnel evacuation, vehicle evacuation, emergency supplies scheduling optimization after accidents, and I solve the model by using relevant algorithm.更多还原