A Risk Assessment Model of Complex Systems Based on Evidential Reasoning Rule with Dependent Evidence

In this paper, an innovative model called evidential reasoning rule with de-pendent evidence (ERR-DE) is proposed for the risk assessment of complex systems under uncertainties. A general risk assessment indicator system is established, and the risk indicators are modelled and described as evidence under the discernment of framework (FoD). The evidence is unified by the transformation matrix, and the evidence reliability and dependence index of evidence are explicitly measured. The ERR-DE model forms a multi-information fusion framework, where multiple pieces of evidence with different weights, reliabilities and dependence indexes are aggregated to establish the relationship between risk indicators and system risk. A parameter optimization model is constructed, where all subjective evidential parameters can be learned through the idea of maximizing expectations. The proposed model is used to assess the risk of a gyroscope.

Keywords: Risk assessment, inferential modelling, evidential reasoning rule, uncertainty, evidence dependence.