SAMPLING-BASED RBDO USING STOCHASTIC SENSITIVITY AND DYNAMIC KRIGING FOR BROADER ARMY APPLICATIONS

摘要: The University of Iowa has successfully developed Reliability-Based Design Optimization (RBDO) method and software tools by utilizing the sensitivity analysis fatigue life; applied to Army ground vehicle components obtain reliable optimum designs with significantly reduced weight improved life. However, this cannot be broader application problems due lack in many areas. Thus, for applications, a sampling-based RBDO using surrogate model been recently. Dynamic Kriging (DKG) is used generate models, stochastic compute sensitivities probabilistic constraints respect independent correlated random variables. Once DKG accurately approximates responses, there no further approximation estimation sensitivities, thus can yield very accurate design. For computational efficiency large-scale engineering problems, parallel computing proposed. Numerical examples verify that proposed finds efficiently.