文献类型：期刊文献

中文题名：基于CU变换的非齐次泊松过程的统计验证模型

英文题名：A Test Model of NHPP: Based on CU Transformation

作者：范朝霞 赵明 杨剑锋

第一作者：范朝霞

机构：[1]贵州大学理学院;[2]耶夫勒大学技术与可持续发展学院;[3]贵州理工学院信息工程学院

第一机构：贵州大学理学院,贵州贵阳550025

年份：2016

卷号：33

期号：6

起止页码：10-13

中文期刊名：贵州大学学报：自然科学版

收录：CSTPCD

基金：贵州省科学技术基金计划(黔科合J字[2015]2064号);高层次人才科研启动经费项目(XJGC20150106)

语种：中文

中文关键词：NHPP检验;CU变换;K-S检验;R语言

外文关键词：NHPP test; CU transformation; K-S test; R language

摘要：开源软件中故障输入数据的非齐次泊松过程(NHPP)的检验和均值函数的估计是排队论分析中重要的子问题。本文依据均值函数的连续性和缓慢变化性,把NHPP转化为小区间泊松性的分段常数非齐次泊松过程(PC NHPP)进行研究,主要运用条件均匀(CU)变换法对数据调整变换后,验证泊松性。本文还运用Bugzilla的错误报告数据对模型实践分析,结果显示CU变换后的验证效果可以接受。最后运用R语言拟合参数,计算出故障输入过程参数,为排队论模型构造和求解奠定基础。
The test of NHPP and the estimation of the mean function about the BUG input data in opened source software are important sub-problems in queuing theory analysis. In this paper,the test of NHPP was converted to the test of Piecewise-Constant NHPP as the continuity and slowness of the mean function,and finally problem converted to the test of the PP on subintervals firstly,and then combined to verify the NHPP on interval. The verification of PP on subinterval is mainly through the CU transformation of the data reasonablely. Also Bugzilla ＇s error report data was used to analyze the model,and the result shows acceptable. Finally,the parameters of the BUG input process are calculated using the R language fitting parameters,which lays the foundation for the construction and solution of the following queuing theory model.