[Under construction]2014 AFOSR MURIA unified mathematical and algorithmic framework for managing multiple information sources of multi-physics systemsPO: Dr. Fariba Fahroo, Computational MathematicsPI: Dr. Karen Willcox, Massachusetts Institute of TechnologyWebsite: TBD
Research problem. Decision processes for complex, multidisciplinary systems draw on multiple information sources. These sources are not commensurable with a scalar-valued measure of “fidelity”—they tell us different things about the problem, with their collective information being greater than the individual parts. We tackle three challenges not addressed by current approaches: (1) managing a range of information sources (multifidelity models, historical, operational and experimental data, expert opinions); (2) certifying analysis and design results; (3) adapting to decision goals. These challenges are further complicated for multi-physics systems, where different information relates to different subsystems and may be coupled in different ways.
Technical Approaches. We propose an integrated research program that leverages the foundations and methods of information theory, decision theory, and machine learning. We bring these elements together in new ways with multidisciplinary and multifidelity optimization, uncertainty quantification, and reduced-order modeling. Our specific research objectives are to: (1) develop statistical approaches for defining and quantifying fidelity; (2) establish decision-theoretic methods for optimally managing sources of uncertain multi-physics information; (3) create reduced-order models (ROMs) with goal-driven adaptation to multi-physics interactions and with quantified uncertainty; (4) formulate an information-theoretic approach for handling multi-physics coupling; (5) create a scalable framework for solving multi-physics analysis and design problems under uncertainty. We propose three research thrusts (RTs): RT1: Optimal information-source management; RT2: Goal-oriented ROMs for the multi-source multi-physics setting; RT3: Managing coupling in multi-physics systems. A specific application of a tailless aircraft integrates the RTs (see Fig. 1).
Anticipated outcome. We will create the first set of principled approaches to analysis and decision-making for multi-physics systems that explicitly integrate the breadth of information sources. Our expected research products are new formulations, methods, and algorithms, demonstrated for our tailless aircraft application system. We will share openly our prototype code implementations and testbed problems.
Potential impact on DOD capabilities. By laying the foundations for multi-source information management, we can impact DoD capabilities through better decisions and improved decision processes. Better decisions can result through a vehicle design process that accounts more fully for uncertainty and for coupling among disciplines, reducing the chances of not meeting requirements, thus reducing expensive re-design work. Improved decision processes will result from a principled approach to harmonization of multiple information sources, and systematic ways to incorporate existing knowledge and expertise. Our application directly relates to Air Force needs and serves as an effective mechanism for building collaborations and achieving transitions with AFRL.