【摘要】 在借鉴海明距离、爬山法思想的基础上 ,提出了通过划分种群 ,采用不同的自适应参数 ,在保持快速收敛于最优的同时不失种群的多样性 .新方法提高了搜索精度 ,较好地克服了未成熟收敛现象 ,并通过多元多峰数学函数对新方法和基本遗传算法进行比较评估 ,验证了新方法具有较满意的优化效果 .更多还原

【Abstract】 On the basis of ideas such as Hamming Distance, Hill Climbing, a metho d is presented to divide the population, use different adaptive parameters, so t hat the requirements of speed and diversity can be met. The new method can enhan ce the searching precision, and effectively overcome the phenomenon of immature convergence. Comparing with the new method and SGA through the mathematical anal ysis, it shows that the new method has obtained a satisfying optimization result .更多还原