The Analysis Of Reliability Growth Model Of Stage Test

With the increasingly wide range and the increasing importance of software applications,the quality of software requirements becomes more and more high. Software reliability is widely recognized as one of the most important aspects of software quality,and it provides quantitative measures of software failures during testing and operating periods.Software reliability growth models(SRGMs) are the primary means of software reliability assessment and prediction.Software reliability growth models(SRGMs)based on non-homogeneous Poisson process(NHPP) have been quite successful tools in practical software reliability engineering.To date,the most appropriate and simplest model to evaluate the customer release reliability is based on the classical NHPP model.Goel-Okumoto model is a classical software reliability growth model based on non-homogeneous Poisson process(NHPP).This model possesses a simple form and is easy to apply,and further more,it works well for plenty of applications.First,in this paper,some classical reliability growth models(Duane model, AMSAA model,J-M model,Musa basic execution time model,G-O model) will be simply introduced and given some important reliability indicators,such as failure rate, reliability function,MTBF.In chapter 3,an in-depth study and discussion is done for G-O model.In addition to the basic reliability indicators mentioned above,appropriate modifications are made for the assumption(3) and the assumption(4) of G-O model,and get two improved models considered the Fault detection rate and the fault removal efficiency respectively.Failure intensity function and reliability function are given.On the basis of chapter 3,in chapter 4,the effectiveness factor and its weighted average are introduced.A reliability growth model,Reliability Growth Model with Immediate Correction and Delayed Correction Mixed,is set up.Reliability of the model is analyzed,and examples show that the new model has better fitting results.