离散结构的遗传形状优化设计

Discreste structural geometrical optimal design using genetic algorithms

离散结构的形状优化问题，设计变量是不同性态的连续／离散混合变量，优化收敛困难。本文利用遗传算法实现其全局最优设计，以空间２５杆桁架结构为例，进行了多变量、多工况的结构形状遗传优化设计。为减少结构重分析次数，引入了Ｓｔｅａｄｙｓｔａｔｅ算法。结果表明结构形状的遗传优化设计方法是可行的。

A new optimization approach—Genetic Algorithms(GA) is adopted to solve the discrete structural geometrical optimal design problem which contains continous and discrete design varibles. A twenty five bar space truss with multiple variables and multiple load conditions is designed by using Genetic Algorithms. Steadystate GA is used to reduce the calculative times of structural reanalysis. The result shows that structural geometrical optimization by using GA is effective.