摘要
遗传算法隐含固有的并行性,并行实现可以加快遗传算法的收敛速度。目前,国内外学者把主要兴趣放在并行机环境中实现机制的研究上,对非并行机环境下的并行实现机制研究较少。文章对标准的遗传算法进行了改进,设计了基于局域网环境的遗传算法的并行实现机制。对100个城市的TSP问题的仿真结果表明,文章设计的并行机制可以提高遗传算法的收敛速度。
Genetic algorithm implies inherent parallel property, and its convergence speed could be improved by parallel realization. At present, the academicians both parallel machine system, while comparatively less on WAN. The standard GA is modified, and the parallel relation method of GA in WAN is designed. A simulation test of 100-city TSP shows that the parallel relation method could improve the convergence speed of G A.
出处
《通信技术》
2009年第5期185-187,共3页
Communications Technology
