摘要
极小子样情况下的抽样仿真一直是可靠性仿真评估的热点和难点,目前没有较好的解决办法。运用灰自助法的思想,改进灰自助法的抽样方法克服自助法在极小子样下的重复抽样问题,运用GM(1,1)二次数据拟合模型克服Bootstrap方法在极小子样下仿真结果不可信问题。将改进的灰自助法运用到极小子样可靠性评估中,多个算例表明该方法仿真评估精度较高,并具有较高的可信度和较强的适用性。
Resampling and simulation with a very small sample is always a hotspot and hard problem in rehability simulation and evaluation. At present, there is no good method suitable to solve this problem. Based on the principle of the grey Bootstrap, the resample method was improved when the Bootstrap repeatly resampled with a very small sample, and GM(1,1) twice data fitting model to solve the problem that the Bootstrap's simulated result is credible with a very small sample. On the improved grey Boot- strap method in very small sample rehability evaluation, the results indicated that the method had high evaluating precision and high credihihty and strong apphcability.
出处
《山东大学学报(工学版)》
CAS
北大核心
2010年第1期144-148,共5页
Journal of Shandong University(Engineering Science)
关键词
改进灰自助法
极小子样
可信度
可靠性
抽样
improved grey Bootstrap
very small sample
credibility
reliability
resample