摘要
服从Weibull分布的寿命试验数据应该是一组随机数,本例研究了它们对信息熵的贡献,并与计算机产生的随机寿命的信息熵进行了对比,结果表明Weibull实测寿命满足随机分布,从而给出可靠的Weibull分布参数κ和λ的充分和必要条件是:全样本极大似然法的κ(Ν)值与从斜度及过盈峭度得出的κ(γ1,Ν)和κ(γ2,Ν)值一致,而且实测寿命全样本的总信息熵趋近最大。
It is believed that life test data should be a random one complied with Weibull distribution. The contribution of Weibull life test data to information entropy is researched, and compared with information entropy of random life pro-duced by computer. The results show that the necessary and complete conditions for a lab - tested life dataset to be a random one and to provide reliable shape parameter and size parameters are： shape parameters calculated respectively by Maximum Likelihood Method, skewness and excess kurtosis are almost overlapped and its total information entropy is approaching maximum at full sample size.
出处
《轴承》
北大核心
2012年第12期28-31,共4页
Bearing
关键词
滚动轴承
疲劳寿命
WEIBULL分布
斜度
峭度
信息熵
rolling bearing
fatigue life
Weibull distribution
skewness
kurtosis
information entropy
