A Marginally Consistent Common-Cause-Failure Bayesian Network Method for Identifying Key Risk Factors in High-Reliability Electronic Circuits
Year: 2026
Type: article
Abstract: High-reliability electronic circuits in safety-constrained equipment often lack sufficient system-level failure data at the early design stage, making risk identification dependent on reliability handbooks, functional structure, and stress conditions. This paper proposes a marginally consistent common-cause failure Bayesian network method that converts handbook-based FIT values into conditional basic-event probabilities and decomposes each basic event into reference and common-cause-induced contributions while preserving its handbook-derived marginal probability. A public electronic safe-and-arm circuit is used for demonstration. Relative to the reference condition, the mobile mechanical, common thermal, and reduced electrical-derating conditions increase the target-event probability by factors of 3.05, 5.77, and 1.02, respectively, while their combined effect reaches 14.54. The thermal condition gives the largest single-condition amplification. Under fixed evidence, the proposed model matches the direct conditional BN, whereas an extra-superposition BN overestimates risk by inflating basic-event marginals. The method supports reproducible common-cause risk inference, key-node ranking, and design-response evaluation for high-reliability electronic circuits.
Source: IEEE Access
Institution: PLA Army Engineering University
Language: English
Subfield: Safety, Risk, Reliability and Quality
Field: Engineering
Domain: Physical Sciences
Open Access status: gold
Estimated APC paid: $2,075
OpenAlex ID: W7202374871