基于遗传算法的多目标可靠性优化

The Multiple Objective Reliability Optimization Based on Genetic Algorithms

在元件的体积、重量和造价的共同约束下的多级串并联系统的可靠性优化问题是一个具有多局部极值的、非线性的、同时具有整数和实数变量的混合优化问题.将遗传算法和多目标可靠性分配问题相结合,对可靠性分配问题进行求解,得到较好效果,从而得出结论,遗传算法在求解多目标可靠性优化问题中是一种行之有效的方法.

A reliability optimization problem of multiple serial-parallel system, with so many constraints such as volume, weight and cost, is a nonlinear optimization problem with a large number of local maximum values. The problem has integer variables such as the number of redundancy component, and also has real variables such as the reliability of the component. Therefore, it is difficult to resolve the complicated problem with traditional methods. In this paper, when using the Genetic Algorithms （GA）, we resolve the problem of reliability apportionment efficiently, and get the better results than before. The results of the calculation manifest that GA is an effective method in solving reliability apportionment problem.