Search

Reliability-based design optimization (RBDO) is a valuable tool for optimizing while considering the impact of uncertainties. However, its application in engineering, specifically in pile foundation design, is complicated due to high computational costs and the potential for nonlinear iteration misconvergence. To address these challenges, we propose a modified optimization calculation method utilizing the mean first-order reliability method (MFORM). The revised function of the reliability index is introduced to ensure computational accuracy and linear regression is employed for its calculation. The results of the case study demonstrate that the modified optimization calculation method not only improves computational efficiency but also enhances computational accuracy. While the form of the performance function significantly influences initial and local optimizations, it has minimal impact on global optimization. Through local and global optimizations, the objective function values are reduced by 20.2% and 24.9%, respectively, for the first form of the performance function. For the second form of the performance function, reductions of 15.0% and 24.9%, respectively, are achieved through local and global optimizations. [ABSTRACT FROM AUTHOR]