考虑故障排除率和引入率的软件可靠性建模  

Software Reliability Modeling Considering Fault Removal Rate and Fault Introduction Rate

作　　者：郝燕玲[1] 陈嵩博[1] 高峰[1] 马林[2] 

机构地区：[1]哈尔滨工程大学自动化学院,黑龙江哈尔滨150001 [2]沈阳炮兵学院电子侦察系,辽宁沈阳110000

出　　处：《计算机仿真》2014年第10期313-316,共4页Computer Simulation

基　　金：中央高校基本科研业务费专项资金支持(HEUCF041410)

摘　　要：软件测试中排错过程的故障排除率并不是100%的,并且由于检测到的尚未被排除的故障影响还会引入新的故障,即存在故障排除率和故障引入率,两者是具有不同的下降趋势的减函数。针对上述问题,提出一种新的非齐次泊松过程软件可靠性增长模型,考虑了随时间变化的故障排除率函数和故障引入率函数。利用一组公开发表的包含故障检测数和故障排除数的软件失效数据集进行仿真与验证,实验结果表明,改进模型具有更好的拟合效果和预测能力。In the correction process the fault removal rate is not 100% when software testing is carried out, and new faults are introduced due to the faults remained, that is, the fault removal rate and the fault introduction rate ex- ist and they decrease with different downtrends. In order to solve this issue, a new software reliability model based on the non - homogeneous Poisson process is proposed, which considers the fault removal rate function and the fault in- troduction rate function which are both time - varying. By validating the model with a set of published software failure data including the numbers of fault detection and fault removal, the proposed model fits the failure data of the soft- ware better and can predict the failure behavior of the software accurately.

关 键 词：软件可靠性增长模型 非齐次泊松过程 故障排除率 故障引入率 软件测试 

分 类 号：N945.12[自然科学总论—系统科学]