Route selection of tank-containing hazardous material transportation in time-varying conditions

The paper is aimed to introduce an innovative method on the optimal choice of hazardous material highway transportation if time choice is allowed to be optional. So far as we know, approximately 80 % of the accidents come up on the highway transportation are said to be related to the transportation of such hazardous materials. In addition, serious accidents are likely to cause heavy casualty and disastrous environmental pollution. To ensure the safety of hazardous material transportation and decrease the damage and loss, the paper has not only extended and improved the dijkstra algorithm, which is widely used to achieve the goal of the static route choice of optimization, but also puts forward an improved method for marking the place names and that of marking the value of the accumulative indexes so as to make the route selection done logically and ideally. Based on the study of the statistical data of more than 500 tank-car transportation accidents, time correction coefficient of such accident rate and leakage probability correction coefficient after the accidents, we have put forward the starting backgrounds for the simulated model establishment. At the same time, a method for marking the accumulative index value has also been used to establish a route selection model under time-varying conditions. In our model, leakage probability on the road sections, the roadside population density and the transport costs are respectively defined as route selection indexes and the principles for route selection. Our case study for such model applications indicates that the model is feasible for such multi-objective optimization issues. It takes into detailed account the most important three route selection factors, that is, the accident probability, the accident severity and the transport cost. Moreover, the proposed model is not only convenient and practical in use but also objective and independent of the subjective influential factors.