Method for reliability analysis based on RBF surrogate model and evidence theory

This paper introduces a research method that assesses the reliability of evidential theory through the application of Radial Basis Function (RBF) agent models. This approach combines the function learning capabilities of RBF agent models with the uncertainty processing strengths of evidential theory to model and predict reliability issues effectively. Initially, RBF agent models are utilized to simulate historical data, creating functional models that form the basis for further analysis. Subsequently, evidential theory is applied to perform a reliability analysis on the predictive results generated by these models. The efficacy and feasibility of this method have been substantiated through real-world case studies, demonstrating its practical application in reliability assessment.