计算机应用 ›› 2005, Vol. 25 ›› Issue (03): 654-656.DOI: 10.3724/SP.J.1087.2005.0654

• 人工智能 • 上一篇    下一篇

基于遗传算法的点群选取模型技术

武占峰,吴裕树   

  1. 北京理工大学计算机科学工程系
  • 发布日期:2005-03-01 出版日期:2005-03-01
  • 基金资助:

    国家部委预研项目

Model of points selection in map using genetic algorithms

WU Zhan-feng,WU Yu-shu   

  1. Department of Computer Science & Engineering, Beijing Institute of Technology
  • Online:2005-03-01 Published:2005-03-01

摘要: 在地图上指定区域内依据一定的约束条件选取一组目标点 (称为点群 ),很多领域都会用到。结合点群目标选取的约束条件和遗传算法的基本原理与特点,设计了一种基于遗传算法的点群目标选取模型。考虑到要最大限度地保持点群的多样性、在内部各地段的分布密度等因素,使用一种家族内相关选择的方法,并提出非固定基因位交叉变异的改进策略。实际计算表明,该算法性能稳定、搜索效率高,节省时间和空间,能有效地避免算法的“早熟”现象,且快速找到全局最优解。

关键词: 点群, 遗传算法, 交叉, 变异, 早熟

Abstract: Points selection in map is used in many fields. Combining the basic principle and characteristics of genetic algorithms with the basic principles of points selection, this paper designed a model of points selection in map based on genetic algorithms. In order to keep the variety of the points, and the internal distribution and density farthest, this paper proposed a new method with Elitist Recombination which employed an improved strategy of the ambulatory loca crossovers and mutations. Experimental results show that the method is timesaving and space-saving,it can avoid prematurity and quickly find optimal solution with stable performance and high search efficiency.

Key words: points, genetic algorithm, crossover, mutation, premature

中图分类号: