Focused on the issue that the sensitivity of hazardous materials transportation routes to uncertain factors is excessively high, a robust vehicle route optimization method for multi-depot hazardous materials transportation was proposed. Firstly, a robust optimization model was designed under the Bertsimas robust discrete optimization theory with the objective function of minimizing transportation risks and minimizing transportation costs. Secondly, on the basis of Strength Pareto Evolutionary Algorithm 2 (SPEA2), a multi-objective genetic algorithm with three-stage encoding was designed for the model. Then, different crossover and mutation operations were performed on the different segments of chromosomes during genetic manipulation,which effectively avoided the generation of infeasible solutions during population evolution. Finally, part of Qingyang Xifeng district road network was chosen as an empirical research example. Distribution plan was carried out at transportation process to form some specific transportation routes. The results show that better robust hazardous materials transportation routes can be quickly obtained by using the robust model and algorithm under multi-depot situation.

Considering the fact that uncertain parameters were widespread in the Traveling Salesman Problem (TSP), this paper built a robust optimization model for traveling salesman problem under the frame of Bertsimas robust discrete optimization theory, and then transformed it into robust counterpart model according to transformational rule. In addition, a single parent genetic algorithm based on Prufer coding was designed to solve the traveling salesman problem. Compared with the traditional genetic algorithm, the method has shortened the length of the chromosome and prevented feasible solutions being destructed by the traditional crossover and mutation operators. According to the validation by numerical examples, the results show that the algorithm has a higher solving efficiency, and the robust model developed in the uncertain environment can get some better robust solutions.