Abstract: After analyzing bi-level characteristics of transportation design, a bi-level transportation planning model with government's expectation of risks minimization and transporter's expectation of costs minimization was developed. With Pydev as its platform, calculation and results could be achieved based on Python programming and TransCAD through genetical?? gorism. Through case study, it was indicated that optimal so lutions could be obtained through generic algorithm and the outcome, close to the lowest network risk was in agreement with expectation.

Key words: hazardous materials transportation, network design, bi-level planning, genetic algorism(GA)

摘要： 分析了危险品道路运输网络设计问题的双层特性,建立了以政府期望的风险最小化为上层目标、运输者期望 的成本最小化为下层目标的危险品运输网络双层规划模型。采用遗传算法,以Pydev为平台,运用Python编程以 及TransCAD生成网络,实现了运算和结果可视化。实例验证结果表明,遗传算法能给出稳定的最优解,而且所得 风险符合预期并接近于最低网络风险。

关键词: 危险品运输, 网络设计, 双层规划, 遗传算法