Hybrid Genetic Algorithm for Multi-objective Vehicle Routing Problem Based on Human-Computer Interaction

The Multi-objective Vehicle Routing Problem (MoVRP) is an important problem in the logistics distribution management, whose two objective functions are to minimize the global transportation cost and to maximize the average customer satisfaction. The mixed integer programming model of MoVRP is proposed in this paper. And a Hybrid Genetic Algorithm based on Human-Computer Interaction (HGAHCI) is presented to deal with this multiple objective problem, which based on collaboration between computer and human. The computer has superior computing speed, while human has different kinds of expertise, for example, human can guide the search directions for finding good solutions. The experiments show that HGAHCI is feasible and effective for solving the MoVRP in real life situations.