有关逆Weibull分布序统计量的一些比较结果.pdfVIP

Aug. 2015 Vol. 32 No. 4 doi: 10.3969/j.issn.1005-3085.2015.04.012 1005-3085(2015)04-0590-09 ∗ QIU Guo-xin , JIANG Hai-bo (1- Department of Science Course, Army Officer Academy of PLA, Hefei 230031; 2- Department of Statistics and Finance, University of Science and Technology of China, Hefei 230026) In this paper, we consider two different systems consisting of independent com- ponents with inverse Weibull distributions whose characteristic life parameters are heterogeneous, but the shape parameters are common. It is shown that the sur- vival function (hazard rate) of a series system is decreasing in the characteristic life parameter vector with respect to p-larger (majorization) ordering. It is also shown that the reversed hazard rate of a parallel system is decreasing (increasing) in characteristic life parameter vector with respect to majorization ordering if the common shape parameters are less (greater) than 1. As a consequence, the simple upper bound on the survival function of the series (parallel) systems is built in terms of geometric (arithmetic) mean of the characteristic life parameters. inverse Weibull distribution; hazard rate ordering; reversed hazard rate ordering; dispersive ordering; majorization ordering; p-larger ordering AMS(2000) 60E05; 62N05 O212.2 A The inverse Weibull distribution is important in reliability, life testing and other engineering applications. In its standard form it has the probability density function f (x) = β (α )exp { − (α ) }, x 0,