Hierarchical C-BRB model based reliability assessment of bearings

Hierarchical Copula and Belief Rule Base (CBRB) model based reliability evaluation of bearings is proposed to address the correlation among indicator factors and the problem of combinatorial explosion. Firstly, the hierarchical reliability assessment model of bearings is established and divided into three sub-rule bases. Then the degree of correspondence between each sub-rule base's indicator factors and corresponding reference value is calculated. Next the correlation coefficient between indicator information is obtained. The parameters of the ideal Copula model are obtained to determine each rule's activation weight. Integrating all the rules through the use of an evidential reasoning (ER) algorithm to get the evaluation results. The hierarchical C-BRB model is optimized using the projection-based covariance matrix adaptive evolution strategy (P-CMA-ES) algorithm to produce the optimal parameters and the final reliability evaluation results. Finally, the model is tested using a bearing dataset.