Software health management with Bayesian networks

摘要

The emerging field of software health management (SWHM) is bound to the critical need to perform tasks such as predicting, detecting, diagnosing, and mitigating adverse events due to software faults and failures. Faults can arise for numerous reasons, including coding errors, unanticipated faults or failures in hardware, or problematic interactions with the external environment. Safety-critical software products, such as aircraft digital controllers, are subject to strict rules of coding, starting with certification, verification, and validation. But software failures will occur anyway, threatening the safety of users.