Unified Reliability Analysis Method for Structures with Multi-source Uncertainties

A unified reliability analysis method is presented to handle the problem of multi-source uncertainties with random, fuzzy and interval variables. In the proposed method, a unified reliability analysis model (URAM) is constructed by applying the conditional failure on the basis of the relationship of probability, possibility, and evidence theories. The URAM belongs to a three layers conditional reliability analysis model, where the basic probability assignment assigned to interval variable is placed in the outermost layer of the conditional failure, the possibility measure assigned to fuzzy variable is placed in the middle layer, and the probability measure assigned to random variable is placed in the innermost layer. An algorithm based on the first order reliability method (FORM) and a-cut, which is referred to as FORM-alpha-URA, is developed to solve the URAM. The feasibility and effectiveness of the presented methodology are demonstrated by a numerical example.