This paper investigates a bi-level optimization for the hazmat transportation problem with lane reservation. The problem lies in how to select lanes to be reserved on the network and plan paths for hazmat transportation tasks. The tradeoff between the transportation cost, the risk and the impact on the normal traffic is considered. Combined with the traffic flow theory, we quantify the impact on the normal traffic and modify the traditional risk measurement model. The problem is formulated as a multi-objective bi-level programming model which involves selecting reserved lanes for government and planning paths for hazmat carriers. Two hybrid metaheuristic algorithms are proposed to solve the bi-level model, which are based on the particle swarm optimization algorithm and genetic algorithm respectively. Their performance on small-scale instances is compared with exact solutions based on enumerating method. Finally, the computational results on large-scale instances are compared and sensitivity analysis on the key parameters is presented. The results show that: 1) Both algorithms are effective methods to solve this problem, and the mothed based on particle swarm optimization algorithm consumes shorter computation time while the mothed based on genetic algorithm has more advantages in optimality; 2) The bi-level model can effectively reduce the total risk of the hazmat transportation while taking into account the interests of hazmat carriers and ordinary travelers; 3) The utilization rate of reserved lanes will increase with the number of tasks. But it is not that the greater the utilization, the better the effect of risk reduction. Once the proportion of hazmat vehicles is excessive, the advantage of reducing the risk of the reserved lane will gradually decrease.

CICTP