对危险货物运输车队进行科学调度，是保障危险货物运输安全和运输效率的一个关键环节。该项目拟在分析危险货物运输事故特点的基础上，综合运用粗糙集和关联规则等理论深入剖析危险货物运输事故机理，挖掘事故致因链；运用贝叶斯网络来预测不同时空条件下的危险货物运输事故率；根据Erkut和Verter提出的“风险公理”构建危险货物车队运输风险模型；然后综合考虑运输风险、公平性、经济性及车队特性等，分别构建静态及动态环境下危险货物运输车队多目标鲁棒调度模型，根据Bertsimas-Sim鲁棒优化理论对鲁棒调度模型进行对等转化，进而构建其高效求解算法；借助ArcGIS平台，依据滚动调度策略构建危险货物运输车队鲁棒调度系统平台，实现危险货物运输车队的动态滚动调度；最后以甘肃省危险货物运输为例展开实证研究，对模型的适用性及解的质量进行评估研究。研究成果可为保障危险货物运输车队安全、经济、高效的调度提供基础理论支持。

The scientific fleet scheduling of hazardous materials is a key point to ensure the transportation safety and transportation efficiency of hazardous materials. This project will intend to make an in-depth analysis of the mechanism of hazardous materials transportation accidents and excavate the causes of accidents by comprehensively applying the theories of rough sets and association rules. Bayesian network is used to predict the accident rate of hazardous materials transportation under different time-space conditions. The fleet transportation risk model of hazardous materials is constructed based on “risk axiom” proposed by Erkut and Verter. Then, the multi-objective fleet robust scheduling models for hazardous materials transportation under static and dynamic environment are constructed respectively by considering the transportation risk, fairness, economy and fleet characteristics. The robust scheduling model is transformed, and then its efficient algorithm is constructed utilizing the Bertsimas-Sim robust optimization theory. Based on ArcGIS, a robust fleet scheduling system platform for hazardous materials is constructed based on rolling scheduling strategy with the dynamic rolling scheduling of hazardous materials fleet. Finally, an empirical research is carried out on the hazardous materials transportation in Gansu Province to evaluate the applicability of the model and the quality of the solution. Our research results can provide basic theoretical support to ensure the safe, economic and efficient scheduling of hazardous materials transportation fleet.