Electromagnetic Risk Analysis for EMI Impact on Functional Safety with Probabilistic Graphical Models and Fuzzy Logic

For the functional safety and security of complex systems such as road vehicles, sufficient immunity of the safety-related electronic, electrical and programmable electronic systems to the anticipated electromagnetic environment is required. Current rule-based practices follow established standards for testing the immunity and emission performance at vehicle and sub-system levels. Nonetheless, with increasingly rapid technological changes, confidence may not be sufficient even if the immunity test levels or the number of tests is increased. A risk-based approach, on the other hand, aims to identify the safety hazards (due to electromagnetic disturbances) and mitigate the associated risks to achieve the required confidence level. However, the traditional tools used for risk analysis (such as fault tree analysis, event tree analysis and failure mode and effect analysis) may not be sufficient as the complexity level increases. In this paper, a graphical approach proposed to enable system visualization as well as supporting a comprehensive risk analysis. The possible implementation of a system-level analysis is illustrated with several methods (e.g. Bayesian networks, Markov random fields and fuzzy set theory), which could be used for estimation of system level.