A New BRB-ER-Based Model for Assessing the Lives of Products Using Both Failure Data and Expert Knowledge

It is vital to assess the lives of newly developed products by using failure data from various testing environments. In the current methods, two steps are generally included. The first step is transforming the failure data under one testing environment into the actual working environment, and the second step is integrating all failure data under the actual working environment into a unified result. However, most available methods cannot use information that includes part failure data and part expert knowledge simultaneously. To resolve the above issue, based on the belief rule base (BRB) and the evidential reasoning (ER) approach, a new BRB-ER-based model is proposed, where the BRB is used to transform the failure data from one testing environment into the actual working environment. The ER approach, which is adopted to aggregate the failure data from different testing environments, is used to assess the life of a product. To conclude, the BRB-ER-based model is applied to represent and integrate asynchronous multisource information. In the proposed model, the initial BRB system is constructed based on experts' knowledge, which results in uncertainty because of the ambiguous nature of human judgment and calls for training the parameters in the BRB-ER-based model. Therefore, an optimal algorithm that employs the differential evolutionary algorithm is proposed. The proposed model and the optimal algorithm operate in an integrated manner to improve the assessment precision by using both failure data and expert knowledge effectively. A case study in three scenarios and use of the conventional approach is examined to demonstrate the capability and potential applications of the new BRB-ER-based model.