Research On The Hybrid Particle Swarm Optimization For Vehicle Routing Problem

Logistics is called "the third profits source", which has been paid more and more attention by people, and increasingly become the foundation of the national economy industry. Distribution is an important link in logistics, accounting for more than60%of the logistics cost. As the core issue of distribution system, vehicle routing problem mainly researches on how to plan the vehicle routes in logistics distribution, in order to improve the economic efficiency, reduce transportation costs, realize scientific logistics.Particle Swarm Optimization (PSO) is a new kind of evolutionary computation and was originally introduced by Eberhart and Kennedy in1995. It has since proven to be a powerful global optimization method, which is been widely applied in combinatorial optimization problem. In this thesis, the logistics distribution as the background, the vehicle routing problem with time windows solved by the particle swarm algorithm is studied on the basis of modern logistics technology.Referring to the Chinese and foreign literatures, according to the basic principles and constraint conditions concerning the vehicle routing problem, the mathematical model of vehicle routing problem with time window is established. Based on the basic principles of particle swarm optimization algorithm, it is established that the particle swarm optimization algorithm the coding method and implementing method to solving the vehicle routing problem. Procedures written by Visual Basic6.0, are used to solve the example concerning vehicle routing problem, in order to proving that the method is suitable for getting the optimal solution by the particle swarm optimization algorithm from the vehicle routing problem. The reasonable parameters setting plays an important role in particle swarm optimization algorithm. What’s more, it is analyzed that the main parameters of particle swarm optimization algorithm affect the performance on solving the vehicle routing problem. Ultimately, how to select the dominating parameters of particle swarm optimization is provided.A hybrid particle swarm optimization algorithm (PSO-SA) is presented to solve the vehicle routing problem, taken into consideration that the primary particle swarm optimization algorithm is apt to fall into the local optimal defect, which is the core of study. According to that particle swarm optimization algorithm is easy to realize with fast convergence rate, and that simulated annealing algorithm has good global search capability, particle swarm optimization algorithm and simulated annealing algorithm are compound, in order to enhance the solution efficiency of the algorithm, and finally achieve the effective improvement on the algorithm, overcoming the defect of particle swarm optimization algorithm, which is ensured that the hybrid particle swarm optimization algorithm is the feasibility and effectiveness of solving the practical problems.The procedures written by VB language are used to realize optimized simulation on the Solomon test data that is the classical vehicle routing problem, and the results about distribution route etc. can be intuitively shown. It is proved that the hybrid particle swarm optimization algorithm is feasible, efficient and universalized to solve the vehicle routing problem, through Satisfactory results that the datas concerning the combined algorithm compared with the particle swarm optimization algorithm.