【作者】 沈小燕；

【导师】 刘浩学；

【作者基本信息】 长安大学 ， 交通运输规划与管理， 2009， 博士

【摘要】 由于危险货物具有易燃、易爆、毒害等特殊理化特性，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，如何确保道路危险货物运输安全，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现实意义。实际证明，道路危险货物运输路线优化是有效降低运输风险的重要措施之一。目前，国内外有关其运输风险分析和路线优化方面的研究，仍然存在一些薄弱点。为此，本文从分析国内外典型事故案例入手，在构建多维度道路危险货物运输风险评价模型的基础上，深入开展多目标道路危险货物运输路线优化研究，主要研究内容如下：①系统分析国内外道路危险货物运输风险分析和路线优化方面的研究现状、发展趋势及存在的主要问题；通过对国内外典型道路危险货物运输事故案例分析，掌握道路危险货物运输事故的发生形态、时间、路段、道路等级以及事故原因等规律，进而确定了道路危险货物运输风险的“5M”影响因素，并针对上述因素中存在的问题提出若干适合我国国情的道路危险货物运输风险减缓措施。②从道路危险货物运输事故概率和事故后果严重度两方面构建道路危险货物运输风险评价模型。在事故概率评价模型中，在确定基于路线区域、道路类型和危险货物种类的道路危险货物运输泄漏事故(包括交通事故引发的泄漏事故和非交通事故引发的泄漏事故)基本概率的基础上，引入事故概率修正系数对其进行修正，构建具有时变特性的道路危险货物运输事故概率。在确定道路危险货物运输事故后果严重度方面，除提出定量化计算影响区域内路上暴露人口、路下暴露人口数量及环境敏感区暴露数量的具体方法外，还利用居民出行比例系数和交通流量变化系数构建了具有时变特性的道路危险货物运输事故后果严重度。③研究了静态条件下“单对单(即单个起点对单个终点)”的多目标道路危险货物运输路线优化问题。在建立相应路线优化数学模型的基础上，针对传统多目标优化模型求解方法中存在主观决策成分过大的缺点，采用灰色关联分析法求解多目标道路危险货物运输最优路线；④探讨了静态条件下“多对多(即多个起点对多个终点)”的多目标道路危险货物运输路线优化问题。针对道路危险货物运输路线选择中面临的双层约束特征，即政府监管部门规划道路危险货物运输网络，危货运输企业在指定运输网络中选择路线这个特征，兼顾政府主管部门和运输企业双方不同利益目标，建立静态“多对多”的双层多目标道路危险货物运输路线优化数学模型，并构造求解该双层路线优化模型的启发式算法。⑤讨论了时变条件下带时间窗限制的“单对单”多目标道路危险货物运输路线优化问题。在建立时变条件下具有允许运输时间窗、允许服务时间窗和允许等待时间窗限制以及多个出发时间的“单对单”道路危险货物运输路线优化的数学模型的基础上，利用时空网络技术将上述模型等效转换为非循环网络中的静态多目标最短路问题，并构造了求解该问题的动态标号规划算法及具体求解步骤。⑥最后，以陕西某危险货物运输企业承运液化石油气运输任务为例，运用所建立的道路危险货物运输风险评价和路线优化模型，在定量化评价LPG运输事故概率以及路上、路下人口暴露风险和环境敏感区风险的基础上，同时考虑运输风险最小化、运输时间(或距离)最小化和运输成本最小化三个路线优化目标，以确定满足上述目标的最佳运输出发时间和具体运输路线。实例研究表明，本文建立道路危险货物运输风险评价模型能够较准确地反应实际运输过程中各路段的风险水平，建立的道路危险货物运输路线优化模型及算法在解决此类复杂性问题上具有良好的可操作性，可为道路危险货物运输主管部门以及运输企业提供路线优化决策技术支持。更多还原

【Abstract】 Because the hazmat have flammable, explosive, poisonous, erosive and other physics and chemistry characteristics, once vehicles carrying these materials take place fire, explosion and other accidents in the transportation process, not only will bring irretrievable damage towards the resident and property, but also will have possibility to cause serious ecological environment pollution. Therefore, how to guarantee hazmat transportation safety and reduce accident risk has vital practical significance to safeguard the people’s personal and property safety and social stability. The reality shows that the route optimization for hazmat transport is one of most effectively measures to reduce transportation risk. But now, the researchs concerning risk analysis and route optimization for road hazmat transport in domestic and foreign still have many weak points. So, this paper begins from the statistic analysis of typical accidents cases, and based on establishing multi-dimension risk assessment model for the road hazmat transport, carries out the research on the multiobjective route optimization of hazmat transport in a deep-going way, the main research aspectsare as follows:①comprehensively reviews the literatures relating the risk assessment and route optimization for the road hazmat transport. Through carrying out the statistical analysis on typical accidents cases of hazmat transport, the characteristics and rules of accidents, including accident pattern, time, and road section and so on are investigated, and then "5M" influencing factors for the road hazmat transport risk is determined. According with the problems that existing in the over-mentioned factors, several risk reduction measures for the road hazmat transportation are proposed, which are suitable for China’s situation.②constructs a multi-dimension risk assessment model for the road hazmat transport from the probability and severity of accident. Among the accident probability model, the basic release probability based on route area, road type and the class of hazmat (involving release probability caused by traffic accidents and by non-traffic accidents) is not only put forwarded, but also the accident probability modification coefficient is introduced to revise the above-mentioned basic probability. Subsequently, the computation method of the time-varying accident probability is proposed. Among the accident consequence severity model, a quantitative calculation method for the en-route and off-route exposed population as well as the exposed environmentally sensitive area in the impact area is put forwarded, and then the time-varying road hazmat transport risk and accident consequence severity is determined by using the proportion of resident trip and the traffic flow variation coefficient.③study on the problem related to multiobjective route optimization for the road hazmat transportation for a given origin and destination under the static condition. Based on establish a mathematical model for the over-mentioned problem, the grey relation analysis method is applied to identify the most satisfied route in all nondominated routes for overcoming subjective factors influence in traditional model.④study on the problem related to the multi-objective rout optimization for the road hazmat transportation for multiple pairs of origin and destination under the static condition. Aimed at two-level restraints in the over-mentioned problem, that is the government designates a subset of the transport network for hazmat transport and carriers select routes on this subset, and given attention to the different benefit targets between government and carriers, a bilevel network flow model for above-mentioned problem is delivered. At same times, a heuristic algorithm that always finds a stable solution for the bilevel model is proposed.⑤discusses the problem related to multi-objective route optimization for the road hazmat transport with time windows constraints and time-varying attributes and for a given pair of origin and destination. Based on establish a mathmatical model which can identify all nondominated time-varying routes with multiple departure time at the origin and waiting time at intermediate nodes along these paths for a given pair of origin and destination, subject to three kinds of practical constraints: limited operational time period, and service, and waiting time window constraints at a node, the proposed model is equivalently transformed into a static multiobjective shortest path problem in an acyclic network reconstructed by the space-time network technique, and an efficient dynamic programming method is then developed.⑥finally, a transport task of the liquified petroleum gas carried out by a transport enterprise of Shaanxi province is taken as an example. On the one hand, the accident probability of LPG transport as well as the exposed risk concerning the en-route and off-route population and the environmentical sensitive area is computed by applying the proposed risk assessment model. On the other hand, by adopting the proposed route optimization model, the most satified LPG transport route with definite departure time at the origin is determined, subject to three optimization objectives: the minimum transportation risk, time (or distance) and cost. The example shows that the proposed risk assessment for the hazmat transport can more accurately reflect the risk level of each road section in actual transport process, and the established route optimization model and algorithm have better operability in solving such complexity problem, which can provide technical support to decision-making of the government authorities and enterprises of the road hazmat transport.更多还原