Resilience evaluation methodology of engineering systems with dynamic-Bayesian-network-based degradation and maintenance

Many resilience evaluation methods were proposed to assess system resilience under the influence of specific, constant disasters, and these methods are restricted by fixed external disasters. A novel evaluation methodology combining Markov model and dynamic Bayesian networks (DBNs) is developed, and is suitable for resilience evaluation under the influence of various external disasters. The external disaster model is established by mapping from actual physical models of influencing factors, the failure rate of a single component is obtained from the established disaster model, the system performance is obtained through the combination of Markov model and DBNs, and the resilience value is finally calculated through the integral of the performance curve. An example of power supply and control systems of subsea blowout preventers demonstrates the application of the evaluation methodology.

Automated summary

A novel evaluation methodology combining Markov model and dynamic Bayesian networks is proposed to assess system resilience under the influence of various external disasters. The method establishes an external disaster model, obtains system performance, and calculates resilience value through the integral of the performance curve. An example of its application in subsea blowout preventers demonstrates its effectiveness.