Research On The Location Routing Problem Model Based On Particle Swarm Optimization

The development of logistics industry is one of the important measures ofcomprehensive national strength and the level of modernization. In the international area,logistics industry is considered to be both the basic industry and promoter of economic innational economic development. With the development of computer information technology，the commodity trading time and space is shortened, the geographical distribution of thelogistics needs is expanding, efficient logistics system is an important factor to improvebusiness efficiency and competitiveness of the industry. Modern logistics system requirescomprehensive consideration of the relevant factors affecting the cost of logistics, at thelowest cost, highest quality of service to meet the needs of each customer. Traditionallogistics system can not meet the modern logistics efficient and fast requirements. So,provides an efficient logistics system to improve the economic efficiency of enterprises,logistics scientific is an important strategy. Location Routing Problem is a vital part oflogistic optimization and must be resolved. In this circumstance, we study the LRP model andit’s algorithm.In the past research, the objectives of LRP is lowest cost, modern logistics not onlyconsider lowest cost, but also the quality of service of the distribution center to improvecustomer satisfaction, customer satisfaction is reflected in the rapid delivery, on the otherway, that is consider customer delivery time limit. Based on the traditional Location RoutingProblem with time windows we analysis the cost function, such as vehicle transport anddistribution costs, the running costs of the distribution centers penalty cost caused bycustomer demand time window and so on. Logistics costs and customer delivery time as theobjective function is optimized, the sum of the cost is minimums, constructed the model ofLocation Routing Problem with Changeable Soft Time Windows (LRPCSTW).Based on the performance of customer satisfaction prevail when in full picking andtimely. The same class model of Location Routing Problem with Changeable Soft TimeWindows (LRPCSTW) is researched. LRPCSTW is a variant of traditional LRP. Thevariable soft time window means that the customer service time limit is a time period, anddifferent customer requirements of the time window is not the same.LRP is a typical NP-hard problem, most of the LRP algorithm based on the idea of thetwo-stage, LRP viewed as two sub-problems: positioning (Location Allocation Problem,LAP), vehicle routing problem (Vehicle Routing Problem, VRP). At first solve thepositioning problem, and then the vehicle routing problem. But the question is sub-problemsoptimal may not be able to guarantee overall optimality, this approach failed to reflect theintegrity of LRP. To this end, on the basis of fully consider the LRP holistic, Particle Swarm Optimization with Swap Operation and Double Mutation algorithm is given to optimize theLRPCSTW with MATLAB (2008a). Finally, the experiments show that the proposedalgorithm is feasibility, stability, effectiveness and robustness. At the same time, given thelogistics system and distribution path visualization. This model and algorithm can be directlyused as a secondary development of GIS logistics system module, a new idea for the design ofthe logistics system based on GIS also provides decision support to the management of theintegrated logistics system.