An Agent-Based Netcentric Framework for Multidisciplinary Problem Solving Environments (Mpse)

Abstract

The process of prot0!JPillg is part of every scientific inquiry, product design, and learning activity. The new economic realities require the rapid prototyping of manufactured artifacts and rapid solutions to problems with numerous interrelated elements. 'This, in turn, requires me fast, accueate simulation of physical processes and design optimization using knowledge and computational models from !l1J(llip/e dircipliJJu (mllllj-p~sj(S OIld om/Ii-Hale IDO&/J) in science and engineering. Thus, the realization of rapid multidisciplinary prototyping is the new grrl/ld cballmge. In iliis application scenario the natural computational resource is a "computational grid" rnat connects the needed distributed hardware and somv:ue resources used w simulate the elements of the artifact. Our research goal is to address this application scenario in Ult: conte.xt of parallel computing, cluster computing (LAN based computational grids), and Intranet/Internet compurational grids. In this document, wc describe the initial design of a generic MPSE framework based on a ne[Work of computational agents assuming a net-centric nm-rimc support environment. Moreover, we prcsent thc realization of this framework for designing a prototype MPSE (GasTurbnLab) for supporting simul:ltions needed for the design of efficient gas turbine engines.

A Two-Grid Method of a Mixed Stokes–Darcy Model for Coupling Fluid Flow with Porous Media Flow

“…The Lagrange multiplier approach is also widely used [14,27] for decoupling multimodel problems. The interface relaxation approach [23,24] has also been successfully applied in multi-model simulations. We note that two-grid methods were proposed in [33,34] for discretizing nonsymmetric and indefinite partial differential equations.…”

Use of Adaptive Metamodeling for Design Optimization

“…Mathematically rigorous techniques and trust regions are also being combined to develop several other metamodel management frameworks to manage the use of approximation models in optimization. [15][16][17][18] Osio and Amon 19 develop a multi-stage kriging strategy to sequentially update and improve the accuracy of surrogate approximations as additional sample points are obtained. The Hierarchical and Interactive Decision Refinement methodology uses statistical regression and other metamodeling techniques to recursively decompose the design space into sub-regions and fit each region with a separate model during design space refinement.…”

Analysis of an interface relaxation method for composite elliptic differential equations

“…Collaborating PDE solvers based on interface relaxation have been already proposed [5,7,9,10,8] and have been rather extensively considered for air pollution [16,17], underwater acoustics [1] and gas turbine engine [6,4] related simulations. In addition, interface relaxation appears to be a very promising framework for coupling Finite Element and Boundary Element methods in electromagnetic applications and beyond (see [18] and references within).…”