Improved Particle Swarm Algorithm And Its Application In Logistics Location

In today’s society, the rapid development of productive,product of social isvery rich. Enterprises’ profits which is obtained by raising labor productivity,riching various products and studying advanced technology is more and morelimited. Therefore, recent years, in order to obtain more profits, enterprises’attention will gradually turned to the problem of how to reduce the cost oflogistics operations. Logistics location in the whole logistics activities played akey role in reduce logistics operation cost. So, by optimization algorithm tooptimize logistics location has a very important research value and significance.Particle Swarm Optimization algorithm (Particle Swarm Optimizati-On, PSO) show the good effect in the continuous optimization problems anddiscrete optimization problem, that due to the factors such as it is simple, easy toachieve, and do not need gradient information, less parameters. In recent years,the PSO algorithm has become a popular research in international in the field ofintelligent optimization, more and more domestic scholars has make the particleswarm optimization algorithm is applied to the peoblem of nonlinearprogramming, vehicle routing, and constrained optimization layout and so on.Aiming at the location problem has been proposed, establish the model, andaccording to the characteristics of the model, solving the model by selecting thesuitable improved particle swarm optimization algorithm, the experimentalresults show the effectiveness of the improved algorithm.The work of this paper as following:1Improving the inertia weight of the particle swarm optimization (PSO)algorithm, putting forward the stochastic inertia weight which is combined withthe search technology of trust region radius, make the algorithm find a balancebetween the ability of global search and the ability of local search, avoidingparticles trapped in local optimum and improving the search efficiency of thealgorithm.2The processing of constraints in optimization problem with constraints,putting forward a improved method which is combined the method of traditional constraints keeping with the search technology of trust region radius,improving the shortcoming that the method of traditional constraints keepingmake the particles easily stagnating, improving the optimization ability of thealgorithm. And calculation by the famous test functions, numerical results showthat the improved algorithm has achieved satisfactory results.3Putting forward improved particle swarm algorithm TRPSO which iscombined with trust region search technique, according to the improved inertiaweight and improved the method of constraint keeping, test by using the testfunctions, and compare the calculated results with the standard particle groupalgorithm and the results of document, it is show that TRPSO algorithm hascertain advantages.4About the constrained optimization problem with narrow feasibleregion, putting forward a improved particle swarm algorithm TMPSO which iscombined with table-manipulation, and analysis of the feasibility of thealgorithm in theoretical.5About location model has the factors of single facility and continuous,using the TRPSO algorithm, and comparing the calculation result with theresults of the method of the center of gravity, and the results of the standardparticle group algorithm, TRPSO algorithm achieved better effect.6About the location model with the factors of multiple facilities and thelocation of the discrete, according to the characteristics of the model, improvingthe model, and solving the improved model by the particle swarm algorithm ofinteger programming, the numerical results show that the improved model andthe algorithm are feasible.7About the location model with the factors of many facilities, continuousand constraint, using the TMPSO algorithm, the experimental results show theeffectiveness of the algorithm.