A Hybrid Particle Swarm Optimization Algorithm for Vehicle Routing Problem with Stochastic Travel Time

Vehicle Routing Problem with stochastic travel time (VRPST) is of crucial importance in today's industries, especially in logistics distribution. This paper presents a hybrid particle swarm optimization algorithm (HPSO) to solve the problem. A chance-constraint model considering capacity of vehicle is founded. The VRPST was changed into a quasi - continuous problem by designing a real number coding. Constrained terms were processed by the penalty function. Cooperating with dynamic neighborhood and the weight value of variable inertia, the proposed HPSO can find the global optimum. The results axe compared with those by both standard particle swarm optimization (SPSO) and improved genetic algorithm (IGA). The illustrations indicate that HPSO can improve success rate of searching best route and is effective for VRPST.