Abstract:
In order to solve the inconsistency of software reliability model and to cast off traditional models’ multi-subjective assumption, this essay adopts genetic programming(GP) evolved algorithm to establish software reliability model based on the software mean time between failure time series. It creates evolution modeling and evaluates its performances index by testing a software from armored force of engineering university. The result certifies the feasibility and the availability of data fitting and forecasting of GP algorithm used on the software reliability modeling, which therefore can be more availably used on software system testing and then secures the availability of data.
Key words:
Software reliability,
Reliability models,
Genetic programming (GP)
摘要: 为解决软件可靠性模型的不一致性,摆脱传统模型多种主观假设的束缚,提出了采用遗传程序设计(GP)的演化算法建立基于软件失效间隔时间序列的软件可靠性模型。针对装甲兵工程学院的某软件测试用例进行演化建模,并对其性能指标进行了分析评价,结果验证了GP算法的可行性以及拟合、预测的有效性,从而能有效地应用于软件系统测试,保障数据的有效性。
关键词:
软件可靠性,
可靠性模型,
遗传程序设计
ZHANG Yongqiang; CHEN Huashan. Software Reliability Model Evolved By Genetic Programming[J]. Computer Engineering, 2007, 33(07): 75-77.
张永强;陈华珊. 软件可靠性的GP演化建模[J]. 计算机工程, 2007, 33(07): 75-77.