Reliability estimation of complex numerical problems using modified conditional expectation method

A simulation-based structural reliability analysis method is presented. It is intended as an alternate approach to estimate reliability for problems for which most-probable point of failure methods fail and when computational resources are limited. The proposed method combines conditional expectation and estimating the PDF or CDF of a selected portion of the limit state. In the proposed approach, complex limit state functions are simplified to two random variable problems. The success of the simplification depends on the quality of the CDF estimate. Results indicate that the method may provide accurate and efficient solutions for some difficult reliability problems.