基于自回归模型的软件可靠性预测方法研究

Research on Software Reliability Prediction Method Based on Autoregressive Model

随着人们对高可靠和高质量软件需求的日益增长,这种需求迫使软件开发人员不断提高软件的可靠性。本文提出了一种基于自回归模型的软件可靠性预测方法,并利用最小二乘估计方法对模型中的参数进行估计。使用开源软件Firefox的真实故障数据,利用SPSS软件对可靠性模型进行验证分析,实验结果表明本文提出的模型是有效的。

With the increasing demand for the people’s pursuit of the high reliability and high-quality software,this demand forces software developers to continuously improve the reliability of software.This paper put for⁃wards a software reliability prediction method based on an autoregressive model,and uses the least square estima⁃tion method to estimate the parameters in the model.The real failure data of the open source software Firefox and SPSS software were used to verify and analyze the reliability model,and the experimental results show that the model proposed in this paper is effective.