Fusion of Modeling and Simulation Credibility in Multidisciplinary Design

Entry vehicle design and aircraft design are just two examples of systems that are of interest to NASA, requiring interactions and exchange of information among multiple performance disciplines. Since any computational optimization process relies on simulation models for identifying the impact of design changes in meeting performance expectations and improving metrics of goodness, it is essential that the uncertainty quantification of these models is captured by the optimization. Fuzzy Logic (FL) provides a systematic approach for introducing linguistic articulation of mental perception into a mathematical framework. In the proposed project the FL approach will be used for introducing in an automated multidisciplinary optimization process the human judgment and the expert opinion associated with the credibility of the modeling and simulations (as stated in the NASA-STD-7009) which are utilized for making decisions. The proposing firm has developed a Decision Support Toolkit (DS Toolkit) which can be used for multidisciplinary design and for balancing many multiple competing performance objectives. The multidisciplinary analysis is done automatically due to specialized algorithms and capabilities which are embedded in the DS Toolkit; both discrete and continuous design variables can be defined. The proposed research will develop the ability to consider the credibility of the models and of the simulations which are used for evaluating the performance requirements and the performance metrics during the analysis. A Fuzzy Logic System (FLS) capability will be developed for this purpose. The membership functions in the FLS will be reflecting the credibility scores assigned by subject matter experts to each one of the eight credibility factors of a simulation. The rule bank in the FLS will capture the expert opinion of the decision makers on how the credibility of the simulations will influences the decisions which are made by the optimization process.