We investigate a hazmat routing-scheduling problem. To minimize the overall expected risk, various vehicles may take different routes/schedules to avoid multiple accidents on the same link. Therefore, the company envisages two issues: (1) unfairly, a vehicle departing earlier from its origin may arrive earlier at its destination than the others leaving later; (2) focusing only on the minimization of risk may increase travel time/cost incurred by the company. We suggest a bi-objective game-theoretic formulation and solve it by a modified Adaptive Large Neighborhood Search and Simulated Annealing. We test the solution on a real-life case and extract practical insights.

我们调查了危险品路由计划问题。为了最大程度地降低总体预期风险，各种车辆可能会采用不同的路线/时间表，以避免在同一条链路上发生多次事故。因此，公司设想了两个问题：（1）不公平地，一辆车辆从其起点出发较早离开可能会比其他车辆晚些到达。（2）仅着眼于降低风险可能会增加公司的旅行时间/成本。我们提出了一种双目标博弈论的表述，并通过改进的自适应大邻域搜索和模拟退火来解决。我们在实际案例中测试该解决方案，并提取实际见解。