Software Reliability Model Considering Fault Severity and Correction Time for HTR-PM Reactor Protection System

The software reliability of digital reactor protection system (RPS) plays an important role in the reliability of RPS. By analyzing the fault data set collected during the software development process of HTR-PM RPS, this paper proposes software reliability models considering fault severity and correction time. The proposed models assume that there are two types of faults with different severity, the ratio of non-critical faults within total faults decays exponentially with time, there is delay between the detection and correction of a fault, and the correction time for faults of the same type is independent and identically distributed. Parameter estimation is carried out by least-square estimation, the mean square error and the coefficients of determination are adopted to evaluate the models. Compared to the models without considering fault severity, the proposed models can reflect the software reliability in more aspects for engineering application.