摘要
针对遗传算法全局优化速度缓慢、搜索的效率对约束惩罚因子的选择有明显的依赖性等问题,介绍了一种能够从可行解空间和不可行解空间同时搜索、具有"精英"保持能力和采用已搜索解集避免了子代的"返祖"和退化现象的快速遗传算法.性能分析表明,该算法为1阶快速收敛的遗传算法,收敛速度优于其它3种算法,而且参数的选择对于算法的收敛速度没有本质的影响,一般在第5次迭代后即可找到全局最优解.
Aimed at problems of genetic algorithm, such as a slower global optimization and search efficiency which obviously depends on the selection of penalty function, a rapid genetic algorithm which can search simultaneously from feasible and infeasible solution space, hold optimal solution and use searched solution space to avoiding reversion and degradation, is introduced . Performance analysis shows that this algorithm is 1 order rapid convergence and convergence rate is better than other three algorithms. Furthermore, the effect of parameters selection has nothing with convergence rate and the algorithm usually finds global optimal solution after fifth iteration.
出处
《云南大学学报（自然科学版）》
CAS
CSCD
北大核心
2009年第5期449-454,共6页
Journal of Yunnan University(Natural Sciences Edition)
基金
国家自然基金资助项目(60572011)
甘肃省自然科学研究基金资助项目(096RJZA115)
关键词
遗传算法
收敛速度
快速
收敛阶
genetic algorithm
convergence rate
fast
convergence order
