可叠加的随机微分方程软件可靠性模型

Additive Stochastic Differential Equations Software Reliability Models

摘要: 随着软件系统规模的不断扩大，大多数复杂软件系统都由多个组件构成，组件的可靠性影响软件系统的可靠性。在实际中，软件故障跟踪系统中故障的报告存在随机性、不规律性以及各种不确定因素，因此故障检测过程可以看作是一个随机过程。本文应用伊藤型随机微分方程建立软件可靠性模型，利用各组件的故障数据来建立叠加的随机微分方程可靠性模型，对模型参数进行估计。最后，将两组实际软件项目的数据集应用于所提出的叠加随机微分方程可靠性模型，结果表明所提出的叠加随机微分方程可靠性模型有更优的效果。

Abstract: With the increasing scale of software systems, most complex software systems are composed of multiple components, and the reliability of the components affects the reliability of the software system. In practice, there are randomness, irregularity and various uncertainties in the reporting of faults in software fault tracking systems, so the fault detection process can be regarded as a sto-chastic process. In this paper, the Itô type stochastic differential equation (SDE) is applied to build a software reliability model, and the fault data of each component is used to build a superimposed stochastic differential equation reliability model and to estimate the model parameters. Finally, data sets from two real software projects are applied to the proposed superposed stochastic differ-ential equation reliability model, and the results show that the proposed superposed stochastic dif-ferential equation reliability model has better results.