An efficient dominant failure modes search strategy and an extended sequential compounding method of system reliability analysis and optimization

Abstract

The structural system is a complex structural form in engineering practice. Under the condition of uncertainty, multiple failure modes may happen simultaneously and lead to the failure of the overall structure. To identify the dominant failure modes of structural systems, this paper proposes a dominant failure mode search method based on the differential evolution algorithm, which generates sample points on the limit state surface and uses the differential evolution algorithm to evolve. It can search for several dominant failure modes of structural systems without much structural analysis function calls. For the evaluation of the reliability of structural systems, this paper further proposes an extended sequential compound method, which can solve the correlation coefficients between compound events and its internal events. Therefore, it can calculate the failure probability for various failure modes and the correlation coefficient between failure modes simultaneously. Besides, a system reliability-based design optimization method is proposed under the framework of the reliability index method with an analytical method for solving the gradient of the system reliability index. Finally, several numerical examples are given to verify the efficiency and accuracy of the proposed methods.