摘要
将基于格雷码的遗传算法用于求解结构的混合离散变量非线性约束优化问题。完整地阐述了混合离散变量的遗传算法处理方法,包括变量的均匀性处理、编码和解码、多余码的处理;采用格雷码编码方式和一种新的适应值定标方法,对遗传算法作了合理的改进,并指出了原编码方式和定标方法存在的问题。并给出算例证明其合理性。
The genetic algorithms, which is based on Gray code, is applied to structure design optimization with mixed discrete variables and nonlinear constraint. A method to deal with mixed discrete variables for genetic algorithms, including how to even variables’values, code and decode and how to deal with superfluous code is presented. An improved genetic algorithms, which is based on Gray code and a new fitness scaling method is developed. We also indicate the disadvantages of natural binary code and usual fitness scaling. Finally, two numerical examples show the efficiency and adaptability of the improved method.
出处
《中国机械工程》
EI
CAS
CSCD
北大核心
1999年第12期1375-1377,共3页
China Mechanical Engineering
基金
国家自然科学基金!资助项目 ( 5 96 85 0 0 3 )
四川省跨世纪杰出青年学科带头人培养基金
关键词
遗传算法
优化设计
离散变量
格雷码
genetic algorithms optimization design discrete variable gray code
