Iteration approach for structural non-probabilistic reliability analysis based on hyper-ellipsoidal models

This paper investigates the iteration algorithms for the non-probabilistic reliability analysis of structures using hyper-ellipsoidal models of uncertainties. For the purpose of alleviating the convergence difficulties of the reliability index calculation in presence of highly nonlinear limit-state functions, the paper proposed a criterion for detecting Zig-zag iterations and a modified iteration scheme based on the HL-RF algorithm. The numerical example demonstrates the validity and efficiency of the present method. The proposed techniques are also shown to be capable of improving the iteration convergence in evaluating the non-probabilistic reliability index.