故障时间观测的机电系统可靠性估计  

Reliability Estimation of Mechanical and Electrical System based on the Observed Failure Time

作　　者：马存宝[1] 李柯[1] 包羽[2] 

机构地区：[1]西北工业大学航空学院,陕西西安710072 [2]中航工业雷达与电子设备研究院第七研究室,江苏无锡214072

出　　处：《机械与电子》2012年第9期3-7,共5页Machinery & Electronics

基　　金：国家自然科学基金资助项目(60572173)

摘　　要：根据系统发生故障时间的统计数据,应用KM方法估计系统可靠度,并利用威布尔分布模型拟合可靠度估计结果。分别用最小二乘和支持向量回归机两种方法估计威布尔分布模型参数,从而得到基于这两种方法的威布尔分布模型。利用所研究的方法,分析了卫星机电系统发生故障时间的统计数据,给出了系统可靠度估计结果,并进行了结果对比分析。According to the statistical data of system failure time,the Weibull distribution model was used to fit the results of the estimated system reliability given by the method of KM. Based on the Weibull parameters estimated by the methods of least squares and Support Vector Regression Machine respectively, the Weibull distribution models were concluded. With the methods described, the statistical data of failure time of the satellite＇s mechanical and electrical system was analyzed. The results of the estimated reliability were offered and analyzed comparatively.

关 键 词：KM方法 可靠度 威布尔分布模型 支持向量回归机 

分 类 号：TH39[机械工程—机械制造及自动化]