Application Research Of Multi Objective Particle Swarm Optimization In Logistics Distribution

Nowadays, the society is a complex system consisted of various networks. It involves multitudinous optimization problems, thereinto, the multi-objective optimization problem is one of the key problems that people are researching at present. As a kind of swarm intelligence algorithm, Particle Swarm Optimization (PSO) finds the optimal solution by means of simulating the biont behaviors of birds foraging and so on, its character of fast convergence rate makes it receive more and more attentions from people. However, Particle Swarm Optimization cannot be applied to multi-objective optimization problems directly, therefore, Multi Objective Particle Swarm Optimization (MOPSO) is developed on the basis of Particle Swarm Optimization. And the textual work was to research the practical application of Multi-objective Particle Swarm Optimization in multi-objective optimization problems.One of the most difficult problems that Multi-objective Particle Swarm Optimization needs to solve is how to determine the global optimum. For the most part, the solutions of the multi-objective optimization problem that the algorithm found are a set of non-dominated solutions called Pareto optimal, rather than a single optimum solution. With the increase of the iterations of particles, the size of the non-dominated solution set will become bigger and bigger, so it is especially important to ensure the algorithm's astringency and distributivity. The Multi Objective Particle Swarm Optimization that this article analyzed uses an external archive to store non-dominated particles and takes advantage of crowding distance to find the global optimum. So that, to a certain degree, it prevents the algorithm falling into local convergence too quickly, and makes the final solutions distribute equably. By means of testing the algorithm by test functions and comparing with other optimization algorithms, the result proves the good performances of the algorithm.Both of the two big problems that the logistics distribution system contains——location problem and routing optimization problem, belong to NP-hard problems. In this article, modified Multi Objective Particle Swarm Optimization was used to settle the practical problems in the logistics distribution system:in the location problem, Discrete Multi Objective Particle Swarm Optimization was used and the location programs were regarded as particles to analyze; while Multi Objective Particle Swarm Optimization baesd on hereditary variation was used and the distribution programs were considered as particles in the routing optimization problem. At last, the algorithm would find the Pareto optimal solutions, sequentially, it proved the problem solving effectiveness of Multi Objective Particle Swarm on such kind multi-objective optimization problems.