基于遗传算法的退化试验优化设计

摘要

在基于退化试验数据的可靠性研究中,如何能在试验资源有限的条件下获得较准确的可靠性评估结果,是退化试验方案设计首要解决的一个问题.以某国产航空液压泵的退化试验方案设计为例,推导并建立了p阶分位寿命分布的方差最小的优化目标函数,以试验预算费用为限定条件,采用遗传算法进行了优化算法设计,并对服从线性退化模型的一批试验样品进行了算例分析.结果表明:与传统优化算法相比,该方法具有运算量小、求解效率高、容易得到全局最优解的特点.%In the research of reliability analysis based on the degradation test data, how to obtain more accurate evaluation results under limited test resources is the topmost problem to solve.A certain kind of domestic aviation hydraulic pump is studied as an example. The optimal objective function with a smallest variance under the 100pth percentile of the lifetime distribution is deduced and established. Then,the genetic algorithm for the optimization is designed,which takes the experimental budget as constraints ,and a group of test samples obeying the linear degradation model is also analyzed. The analysis result shows that: compared with the traditional optimization algorithms,the discussed method has the characteristics of small computation amount and high solving efficiency,and is easy to obtain the global optimal goal.