摘要
针对竞争失效产品加速寿命试验存在试验时间长、费用高、效率低的问题,提出了一种基于Monte-Carlo仿真的竞争失效产品加速寿命试验优化设计方法。采用Monte-Carlo对竞争失效产品的加速寿命试验进行仿真模拟,以正常使用应力下的p阶分位寿命渐近方差估计最小为目标,以各试验应力水平及对应应力下的试验截尾数作为设计变量,采用MLE理论进行统计分析,建立了基于仿真的竞争失效产品加速寿命试验优化设计模型。最后通过GA-BP神经网络对目标函数进行拟合,降低了仿真规模,提高了试验效率,为电子装备寿命预测的加速试验方案优化设计提供技术支撑。
Considering the problems of long test time, high cost and low efficiency of the accelerated life test of products with competing risks, we proposed a method of Monte-Carlo simulation based optimal designs for the accelerated life test. The accelerated life test was simulated with Monte-Carlo method, taking the asymptotic variance estimation of 100 pth percentile of the lifetime distribution of the product being minimum as the objective, and taking each test stress level and the test censored data of the corresponding stress as the design variables. Statistics analysis was made by using MLE theory, and the optimal design model of accelerated life test with competing risks was established based on the simulation. The objective function was fitted by using GA-BP neural network, which reduced the emulation scale and improved the efficiency of the test. The design may provide a technical support for optimal design of accelerated life test in life prediction of electronic equipment.
出处
《电光与控制》
北大核心
2013年第8期95-99,共5页
Electronics Optics & Control
基金
河北省重点基础研究项目(109635529D)
