摘要
针对机械系统可靠性试验数据分析中小子样疲劳性能数据难以满足经典统计方法容量要求的问题,将Bootstrap方法引入到小子样下母体百分位置信下限的计算。提出一种半参数Bootstrap方法用于计算正态分布变量的可靠度置信下限,并以某数控装备伺服驱动系统的可靠性分析实例验证该方法的可行性。
The middle or small samples of fatigue performance data from mechanical system reliability analysis experiment can not meet the requirements of capacity in traditional statistical method. The Bootstrap method is adopted to calculate the matrix-percentile lower limitation of confidence level. A semi-parameter Bootstrap method is proposed to calculate the reliability lower limitation of confidence level for normal distribution variables. An example of reliability analysis for the servo drive system of CNC equipment is introduced to validate the feasibility of the methods.
出处
《机械设计与制造》
北大核心
2010年第3期105-107,共3页
Machinery Design & Manufacture
基金
国家863高技术研究发展计划资助项目(2006AA04Z407)
国家自然科学基金(50705032)
