A Hybrid Genetic Algorithm for the Periodic Vehicle Routing Problem with Time Windows

We propose a new population-based hybrid meta-heuristic for the periodic vehicle routing problem with time windows. Two neighborhood-based meta-heuristics are used to educate the offspring generated by a new crossover operator to enhance the solution quality. This hybridization provides the means to combine the exploration capabilities of population-based methods and the systematic, sometimes aggressive search capabilities of neighborhood-based methods, as well as their proficiency to explore the infeasible part of the search space to both repair infeasible solutions and tunnel toward, hopefully, improved ones. Extensive numerical experiments and comparisons with all methods proposed in the literature show that the proposed methodology yields very high quality solutions, improving those currently published.