Reliability and risk assessment of digital twin system based on improved failure mode and effects analysis

Digital twin system (DTS) is now widely employed in all spheres of industrial production, and the study of its risk and reliability is essential for the stable operation of the real and virtual models to function as they should. Therefore, the reliability and risk assessment of a specific digital twin system based on the intelligent gas application are the focus of this article. Due to the possibility of interaction between the potential failure modes of the DTS, in this paper, the most commonly used failure mode and effects analysis (FMEA) model for industrial applications are improved to analyze the DTS. We employ the trapezoidal fuzzy theory, the triangle fuzzy analytic hierarchy process and the standard deviation approach to reduce the inherent flaws of traditional FMEA such as too much subjectivity and ranking values that do not take into account risk factor weights, and the TODIM (an acronym in Portuguese for interactive and multiple attribute decision-making) which modified by entropy value to consider the interrelationship between failure modes. Finally, we analyze with the aid of the Bisecting K-means clustering and the ranking result under the proposed model is compared with those other three models, and lastly, the proposed model is validated to be practical to assess the risk and reliability of DTS.