基于风险分析的危险品道路运输路径选择模型及算法研究

发布时间：2018-07-17 14:47

【摘要】：由于我国铁路、管道、水路等交通部门对危险品运输尚未提出有效的管理办法，且设施条件相对道路运输落后并存在缺陷，加之道路运输便于与其他运输方式衔接，可快速实现“门对门”运输，所以危险品道路运量逐年提高。危险品的物理化学性质导致其与普通货物不同，若其在运输途中发生事故，不仅严重地威胁沿线居民人员的正常生活，带来巨大的经济损失，而且对周围的环境、建筑设施等造成不可恢复的污染和破坏。因此危险品道路运输备受关注。选择一条合适的运输路径，确保危险品安全经济的运输到目的地对社会稳定可以发挥重要的作用。危险品运输风险分析是路径选择的基础。本论文首先归纳总结了国内外学者提出的风险模型，分析了事故发生概率、事故影响后果以及应急响应能力对运输风险的影响，得出了有效的应急响应对风险分析的重要性，在传统风险模型中引入有效应急响应系数，修正了模型。然后在风险分析的基础上，对静态条件下的危险品道路运输路径选择进行研究。假定运输风险和运输成本在各路段是一定值，与时间无关。当路网中仅有一种危险品运输时，仅从运输商角度追求效益和安全双优，建立运输风险最小和运输成本最少的双目标模型，并利用遗传算法求解；在现实运输网络中，可能存在两种及其以上的危险品同时运输，为了保证整个路网的稳定，需要政府部门和运输商双层约束，鉴于此，建立了上层以政府部门期望的总运输风险最小且风险均衡，下层以运输商期望的总运输成本最少的双层数学模型，并利用启发式算法对模型求解。最后考虑了时间因素对运输风险和运输成本的影响，同时，为了降低潜在发生事故风险，需要限制危险品车辆在规定的时间段内通过某路段，基于此，建立了时变条件下基于运输风险最小和运输成本最少且有时间窗约束的双目标数学模型。针对有软时间窗约束的问题，若车辆经过某路段违反时间窗时，用增加目标值的方法来惩罚，，而对于有硬时间窗约束的问题，禁止车辆违反时间窗，否则被视为不可行解。对于提出的模型，利用动态规划和标号法相结合的算法求解。本论文的研究结论可为危险品在实际路网中兼顾成本与安全的路径选择提供理论依据。
[Abstract]:Because the railway, pipeline, waterway and other transportation departments of our country have not yet put forward effective management methods for the transportation of dangerous goods, and the facilities are backward and defective relative to the road transportation, and the road transportation is convenient to link up with other modes of transportation, Can quickly realize "door-to-door" transport, so dangerous goods road traffic increased year by year. The physical and chemical properties of dangerous goods make them different from ordinary goods. If accidents occur in the course of transportation, they not only seriously threaten the normal life of residents along the route, but also bring huge economic losses to the surrounding environment. Unrecoverable pollution and damage caused by construction facilities. As a result, road transport of dangerous goods has attracted much attention. Choosing a suitable route to ensure the safe and economical transport of dangerous goods to the destination can play an important role in social stability. Risk analysis of dangerous goods transportation is the basis of route selection. This paper first summarizes the risk model put forward by domestic and foreign scholars, analyzes the probability of accidents, the consequences of accidents and the impact of emergency response capacity on transport risks, and concludes the importance of effective emergency response to risk analysis. The effective emergency response coefficient is introduced into the traditional risk model, and the model is modified. Then, on the basis of risk analysis, the road transportation route selection of dangerous goods under static condition is studied. It is assumed that transport risk and transportation cost are certain value in each section, independent of time. When there is only one kind of dangerous goods transportation in the road network, the double objective model of minimum transportation risk and the least transportation cost is established only from the perspective of the transporter, and the genetic algorithm is used to solve the problem. There may be two or more kinds of dangerous goods transported at the same time. In order to ensure the stability of the whole road network, both government departments and transporters are required to restrain each other. In view of this, a minimum and balanced total transport risk expected by government departments is established at the upper level. The lower layer is a bilevel mathematical model with the least total transportation cost expected by the transporter, and the heuristic algorithm is used to solve the model. Finally, the influence of time factors on transportation risk and transportation cost is considered. In order to reduce the potential accident risk, it is necessary to restrict dangerous goods vehicles from passing through a section of road within a specified time period. A two-objective mathematical model based on the minimum transport risk and the least transportation cost with time window constraints is established under time-varying conditions. For the problem of soft time window constraint, if the vehicle passes through a section of the road and violates the time window, the method of increasing the target value is used to punish the problem, but for the problem with hard time window constraint, the vehicle is prohibited from violating the time window, otherwise it is regarded as an infeasible solution. For the proposed model, the dynamic programming and labeling algorithm are used to solve the problem. The conclusion of this paper can provide a theoretical basis for the route selection of dangerous goods in the actual road network, which takes both cost and safety into account.
【学位授予单位】：兰州交通大学
【学位级别】：硕士
【学位授予年份】：2014
【分类号】：U492.336.3

【参考文献】