Nuclear safety-critical Digital Instrumentation and Control system software: Reliability demonstration

Abstract A software reliability demonstration method was designed for the modeling and analysis of nuclear safety-critical Digital Instrumentation and Control (DI&C) system software. A “Reliability Demonstration of Safety-critical Software, RDSS” platform based on the Claim-Argument-Evidence (CAE) is presented to help the analysts for the analysis and evaluation of the software reliability. A novel method transformed the CAE method into Bayesian Brief Network (BBN) was proposed for the mining of the most sensitive elements (Evidence) in the DI&C system by sensitivity analysis with three reasoning methods. Moreover, this paper proposes an engineering approach for the identification of the vulnerability elements which influenced the reliability of DI&C system software life cycle.