A hybrid multi-objective evolutionary optimization approach for the robust vehicle routing problem

In this paper, we propose to approximately solve the robust vehicle routing problem with a population-based method. Uncertainty can be modeled by a set of scenarios where each scenario may represent the travel costs assigned to all visited arcs of the graph associated to the problem. Unlike several existing methods that often aggregate multiple objectives into a compromise function, the goal of the proposed approach is to simultaneously optimize both the number of vehicles to use and the worst total travel cost needed. The proposed method can be viewed as a new version of an evolutionary approach which is reinforced with a “strong-diversification”. Such a strategy is based upon destroying and re-building procedures that are hybridized with a local search using a series of move operators. A number of experiments have been conducted to assess the performance of the proposed approach. Its achieved results have been tested on benchmark instances extracted from the literature and compared to those reached by the-state-of-the-art GLPK solver and one of the most recent method available in the literature. The proposed method remains competitive, where encouraging results have been obtained.