为了搜索函数最优解,基于遗传算法基本理论,提出了良性进化的自适应遗传算法(AGA)。AGA 从两个方面改进了标准遗传算法:一是交叉、变异率会自适应调节大小;二是交叉、变异具有方向性。通过对 AGA 的仿真研究,分析了 AGA 中参数取值对算法的性能影响。最后把 AGA 和标准遗传算法进行了仿真比较,结果表明 AGA 在求解函数最优解问题时具有较强的自适应性和收敛性。
[Objective] The research aimed to study the significant influence factors of the population variations of oriental fruit fly. [Method] Using stepwise regression analysis, the population variations law of oriental fruit fly in Jianshui County of Yunnan province and the meteorological factors that caused its occurrence were analyzed. And the regression model was built. Finally, the regression model was tested on the basis of the data in Jianshui County of Yunnan Province during 2004-2006.[Result] The main meteorological factors that influenced the occurrence of oriental fruit fly were relative humidity, the lowest monthly temperature and rainfall. [Conclusion] This study will provide certain reference for the prediction researches on the time, quantity and occurrence peak of oriental fruit fly.
