不完美排错软件可靠性增长模型效用量化研究  

Research on utility quantification of reliability growth model for imperfect debugging software

作　　者：张策[1] 孙智超 王金勇 袁雨飞 盛晟 吕为工 Zhang Ce;Sun Zhichao;Wang Jinyong;Yuan Yufei;Sheng Sheng;Lyu Weigong(School of Computer Science&Technology,Harbin Institute of Technology,Weihai Shandong 264209,China;School of Automation&Software Engineering,Shanxi University,Taiyuan 030006,China;Dept.of Computer Science,University of Copenhagen,Copenhagen 21004,Denmark;Shenzhen Huantai Technology Co.,Ltd.,Shenzhen Guangdong 518063,China)

机构地区：[1]哈尔滨工业大学(威海)计算机科学与技术学院,山东威海264209 [2]山西大学自动化与软件学院,太原030006 [3]哥本哈根大学计算机科学系,丹麦哥本哈根21004 [4]深圳市欢太科技有限公司,广东深圳518063

出　　处：《计算机应用研究》2022年第12期3724-3729,共6页Application Research of Computers

基　　金：国家自然科学基金资助项目(61473097);山东省自然科学基金面上项目(ZR2021MF067);山西省基础研究计划资助项目(201801D121120);威海市科技发展计划资助项目(ITEAZMZ001807)。

摘　　要：为了进一步提升现有非齐次泊松过程类软件可靠性增长模型的拟合和预测性能,首先从故障总数增长趋势角度对不完美排错模型进行深入研究,提出两个一般性不完美排错框架模型,分别考虑了总故障数量函数与累计检测故障函数间的线性关系与微分关系,并求得累计检测的故障数量与软件中总故障数量函数表达式;其次,在六组真实的失效数据集上对比了提出的两种一般性不完美排错模型和六种不完美排错模型拟合预测性能表现。实例验证结果表明,提出的一般性不完美排错框架模型在大多数失效数据集上都具有优秀的拟合和预测性能,证明了新建模型的有效性和实用性;通过对提出的模型与其他不完美排错模型在数据集上的性能的深入分析,为实际应用中不完美排错模型的选择提出了建议。In order to further improve the fitting and prediction performance of the existing non-homogeneous Poisson process class software reliability growth models,this paper first conducted an in-depth study of imperfect fault exclusion models from the perspective of the total number of faults growth trend.Considering the linear and differential relationships between the total number of faults function and the cumulative detected faults function,it proposed two general imperfect fault exclusion framework models,and derived the expressions of the cumulative detected faults and the total number of faults function in software.Secondly,this paper compared the performance of the proposed two general imperfect debugging models and the six imperfect debugging models on six real failure datasets for fitting prediction performance.The results of the case validation show that the proposed general imperfect debugging framework model has excellent fitting and prediction performance on most of the failure datasets,which proves the effectiveness and practicality of the new modeling framework.Again,through the in-depth analysis of the performance of the proposed model and other imperfect debugging models on the datasets this paper puts forward some suggestions for the selection of imperfect debugging models in practical applications.

关 键 词：非齐次泊松过程 软件可靠性增长模型 不完美排错 效用分析 

分 类 号：TP311.53[自动化与计算机技术—计算机软件与理论]