基于改进遗传算法的离散公差模糊稳健优化设计

为了解决离散变量优化设计问题，采用了一种面向设计、制造的设计变量离散化处理方法。通过改进最优个体保护策略和混沌移民算子，并引入家族竞争遗传机制对基本遗传算法进行了改进。开发了混合离散变量的改进遗传算法程序LSGA2005，用该程序解决了箱体零件尺寸链的离散公差模糊稳健设计问题。实例表明，该算法全局收敛能力强，具有较好的实用价值。

Fuzzy robust optimization design of discrete tolerance based on improved genetic algorithm

For the sake of solving the optimization designing problems of discrete variables, a kind of discretized processing method of designing variables oriented to design and manufacturing has been adopted. By means of improving the protection strategy of optimal individual and the chaos immigration operator and introducing the genetic mechanism of family competition to carry out improvement on the basic genetic algorithm. The improved genetic algorithmic program LSGA2005 of the mixed discrete variables was developed. By the use of this program, the fuzzy robust designing problem of discrete tolerance for the dimension chain of casing part was resolved. The living example indicated that this algorithm is strong in the overall convergent ability and has rather Rood practical value.