Abstract: A novel method based on sequence Shepard interpolation is presented for reliability analysis of structure with implicit limit state equation. Using experiment points of uniform distribution generated by hammersly point set, the response values of samples from importance sampling probability density function are predicted by Shepard interpolation method, and the failure probability of structure is estimated simultaneously. In the process of importance sampling, the most probable failure point searched in current iteration is used as the center of the experiment point set and the density center of the importance sampling probability density function in the next iteration. Comparing with Monte Carlo method, three numerical examples show that the sequence Shepard interpolation is more efficient. The reliability analysis of flap mechanism verifies the engineering applicability of this method.

Key words: Failure probability, Hammersly point set, Importance sampling method, Monte-Carlo method, Shepard interpolation