Research And Application Of The GIS Optimal Location And Its Parallelization Based On Artificial Bee Colony Algorithm

With the development of society and city, the conflict of the irrational distribution of public service facilities and needs of urban residents become more and more prominent. The position quality is direct impact on whether the public services are able to maximize its effect. Optimizing space involves complex nonlinear optimization problem, it runs slow, if using the traditional method to calculate. So we need to find an effective way to solve the problem.In this paper, according to the characteristics of the spatial geographic information, using artificial colony algorithm, we realized the basic artificial colony algorithm in the optimum location of space, and combined with CUDA parallel architecture and algorithm of parallel mechanism, realized the parallel artificial colony algorithm in the optimization space location.First of all, the paper analyzes the factors influencing the location of the space optimization, considered capacity, regional characteristics, scope. We put the related factors in the quantitative, the quantitative factors into the mathematical model, the target fitness function is determined.Secondly, on the basis of our analysis of the colony algorithm and the exchange mechanism between the colonies, the artificial bee colony algorithm was proposed. At the same time, based on the characteristics of spatial geographic information, aiming at the shortcomings of the artificial colony algorithm, we introduced the cross idea into the algorithm, proposed enhanced artificial swarm algorithm(EABC) in the location of the space optimization. Space optimization location belongs to nonlinear combinatorial optimization problem, large amount of calculation and running time is long, so this paper is based on CUDA parallel architecture, parallel artificial colony algorithm was proposed, and the algorithm model has carried on the related design.Finally, we designed and implemented the serial and parallel artificial colony algorithm in the application of spatial optimal location, analyzed the effect of location. And we compared the serial and parallel algorithms from two aspects of running time and accuracy. Experiments show that the artificial colony algorithm in the optimal location of space shows its superiority, parallel artificial colony algorithm achieved good speedup.