Sequential optimization and possibility assessment, a new efficient method for reliable optimal design

摘要

In this paper, the "Reliability-Based Design Optimization (RBDO)" is used to solve an uncertain design optimization problem with insufficient experimental data. For modeling the uncertainty in these cases, the possibility theory is applied and a new method in the "Possibility-Based Design Optimization (PBDO)" strategy is introduced. The efficiency of the proposed approach is compared with the basic PBDO method (PMA) using two engineering examples. The two facts that in such cases "the PBDO could provide conservative solutions" and "satisfy system level reliability" but RBDO couldn't, are studied too.