改进的混沌遗传算法及结构优化设计

Improved Chaos GA and Structural Optimum Design

提出了一种结构新颖、快速收敛的混沌遗传算法,它将交叉和变异遗传运算融为一体,通过调节变异调整系数使生成的点在上一代最优点的附近随机摆动,避免了随机漫游现象。每代仅保留上一代的最优个体,其它个体重新生成,在最优点附近利用混沌变量的遍历性进行局部寻优,提高了解的精度。改进的混沌遗传算法可以克服普通遗传算法经常出现的未成熟收敛、振荡以及随机性太大的问题。改进的混沌遗传算法是一种有效的优化方法,适合于大型复杂结构优化设计。

A kind of novel fast convergence GA is studied. It makes the new points swing around the best point of the last generation and can avoid the random roam. Except keeping the best individual of the last generation, all the other individuals are reproduced. Searching at the best point＇s neighborhood using chaos variables can ensure the solution＇ s precision. Improved chaos GA can overcome prematuration swing and stochastic etc.. Solved the optimization of test functions via Improved chaos GA and SGA respectively. It is concluded that Improved chaos GA is more effective. Improved chaos GA is an effective optimization for optimum design of large - scale structure.