Adaptive Importance Sampling Algorithm for Fuzzy-Random failure Probability

An adaptive importance sampling algorithm is presented to evaluate fuzzy-random general failure probability,where a simulated annealing optimization is employed to seek the fuzzy-random design point—the Most Probable Failure Point(MPFP) in the failure region.The efficiency of the simulated annealing is more than that of the first order and second moment for seeking the MPFP,this method is especially suitable for the failure probability computation with the highly non-linear failure region and the complicated equivalent joint probability density function.The contribution of the set of importance sampling functions to the fuzzy-random failure probability is represented by the designed simulation number to lead to more accurate computation.The illustrations verify the feasibility and efficiency of the proposed method.