一种改进的自适应遗传算法及其在层合板优化中的应用研究

摘要

In order to overcome the shortages of prematurity and worse convergency for conventional genetic algorithm,a new strategy for improving efficiency of the conventional genetic algorithm has been put forward.In the present method,the selection probability and the mutation probability are self-adaptive.Taking the stability of laminate composite as the goal,the optimized laminate sequency is analyzed with the help of using the benchwork of ABAQUS,and the new GA program,the parametric modeling program and the eigenvalue searching program are coded with Python language.Comparing with the conventional genetic algorithm,the new GA method shows a good consistency,fast convergency and practical feasibility.%针对传统遗传算法早熟与收敛性差的问题对遗传机制提出了改进策略,实现了选择概率和变异概率的自适应。研究了在ABAQUS平台下对层合板实施以稳定性为优化目标的优化设计技术,使用Python语言编写了改进后遗传算法的优化程序、ABAQUS参数化建模程序和有限元分析后结果文件搜索程序,通过算例计算得到了层合板实现最大特征值的优化铺层顺序,结果表明与简单遗传算法相比,改进后的方法最优结果一致性较强,收敛速度较快,可行性高。