A STUDY ON THE QUANTITATIVE RELIABILITY ESTIMATION OFSAFETY-CRITICAL SOFTWARE FOR PROBABILISTIC SAFETYASSESSMENT

摘要

Recently practical needs have required quantitative features for the softwarernreliability for Probabilistic Safety Assessment (PSA) which is one of the importantrnmethods being used in assessing the overall safety of nuclear power plant (NPP). But thernconventional assessment methods of software reliability could not provide enoughrninformation for PSA of NPP, therefore current assessments of a digital system whichrnincludes safety-critical software usually exclude the software part or use arbitrary values.rnThis paper describes a Bayesian Belief Networks (BBN) based method that models thernrule-based qualitative software assessment method for a practical use and can producernquantitative results for PSA.