This proposal seeks to pioneer innovative methods for managing data on various levels of fidelity through extensions of previous methods, computational results, and rigorous mathematical results. Specifically, Multifidelity Sequential Kriging Optimization (MFSKO) will be extended to address multicriteria optimization involving more than a single type of model representing more than a single discipline. Also, rigorous convergence results from Schonlau (1997) will be extended to multifidelity optimization in the context of radial basis function methods and Kriging models. Adaptive methods will be developed to achieve probabilistic convergence results and enhance performance. Results will be illustrated using a flying wing UAV design function integrating information from structural and fluid models. BENEFIT
Keywords: Multiple Fidelity, Global Optimization, Surrogate Systems, Meta Models, Hierarchical Systems, Uncertainty Management
