Development of Multidisciplinary, Multifidelity Analysis and Integration of Aerospace Vehicles

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: Constructing a broad, yet detailed and accurate, portrayal of the design space will help designers manage the design risk and expand design options while remaining cognizant of the associated uncertainties and of the remedies/options available to reduce these uncertainties.  Use of the proposed technology will help reduce the design and life-cycle cost of next-generation high-efficiency flight vehicle systems, and will help attain better designs, by providing a better understanding of how the design variables interact and influence each other under the influence of uncertainty and by incorporating these interactions early in the design process.  Military applications include the design of advanced sensor platforms, UCAVs, and space launch vehicles.  A diverse range of design applications exists beyond the aerospace field, including for example in the renewable energy, chemical and metallurgical industries, in macroeconomics and pharmaceuticals, and in the optimization of manufacturing processes.)