다분야 통합 최적설계에서 근사모델 관리기법의 활용
Utilization of Application Model Management Techniquesin Multidisciplinary Design Optimization
0
조회
0
다운로드
1
인용
0
참고문헌
키워드
초록
In system design, it is not always possible that all decision makers can cooperate fully and thus avoid conflict. They each control a specified subset of design variables and seek to minimize their own cost functions subject to their individual constraints. However, a system management team makes every effort to coordinate multiple disciplines and overcome such noncooperative environment. Although full cooperation is difficult to achieve, noncooperation also should be avoided as possible. Our approach is to predict the results of their cooperation and generate approximate Pareto set for their multiple objectives. The Pareto set can be obtained according to the degree of one's conceding coupling variables in the other's favor. We employ approximation concept for modelling this coordination and the mutiobjective genetic algorithm for exploring the coupling variable space for obtaining an approximate Pareto set. The approximation management concept is also used for improving the accuracy of the Pareto set. The exploration for the coupling variable space is more efficient because of its smaller dimension than the design variable space. Also, our approach doesn't force the disciplines to change their own way of running analysis and synthesis tools. Since the decision making process is not sequential, the required time can be reduced comparing to the existing multidisciplinary design optimization. This approach is applied to some mathematical examples and structural optimization problems.
Abstract
In system design, it is not always possible that all decision makers can cooperate fully and thus avoid conflict. They each control a specified subset of design variables and seek to minimize their own cost functions subject to their individual constraints. However, a system management team makes every effort to coordinate multiple disciplines and overcome such noncooperative environment. Although full cooperation is difficult to achieve, noncooperation also should be avoided as possible. Our approach is to predict the results of their cooperation and generate approximate Pareto set for their multiple objectives. The Pareto set can be obtained according to the degree of one's conceding coupling variables in the other's favor. We employ approximation concept for modelling this coordination and the mutiobjective genetic algorithm for exploring the coupling variable space for obtaining an approximate Pareto set. The approximation management concept is also used for improving the accuracy of the Pareto set. The exploration for the coupling variable space is more efficient because of its smaller dimension than the design variable space. Also, our approach doesn't force the disciplines to change their own way of running analysis and synthesis tools. Since the decision making process is not sequential, the required time can be reduced comparing to the existing multidisciplinary design optimization. This approach is applied to some mathematical examples and structural optimization problems.
참고문헌 (13)
- S. Burgee, "A coarse grained parallel variable-complexity multidisciplinary optimization paradigm", The International Journal of Supercomputing Applications and High Performance Computing, 10, pp.269~299 (1996)
- A. A. Giunta, "Aircraft Multidisciplinary Design Optimization Using Design of Experiments Theory and Response Surface Modeling Methods", Phd, Department of Aerospace and Ocean Engineering, pp.1997
- J. Rodriguez, "Trust region augmented Lagrangian methods for sequential response surface approximation and optimization", Proceedings of DETC '97 (1997)
- A. J. Booker, "A rigorous .framework for optimization of expensive functions by surrogates", Structural Optimization, 17, pp.1~13 (1999)
- J. E. Dennis Jr, "multidisciplinary Design optimization: State-of-the-Art", pp.330~347 (1997), Department of Computational and Applied mathematics, Rice University
- Numerical optimization using computer experiments, "M. W. Tosset", ICASE, NASA Langly Research Center, ~, pp.Technical Report 97-38 (1997)
- Timothy W. Simpson, "A Concept Exploration Method for Product Family Design", Phd, Department of Mechanical Engineering, pp.1998
- S. H. Nasseri, "Optimized solution of a two-bar truss nonlinear problem using fuzzy Geometric programming", Journal of Nonlinear Analysis and Application, 2014, pp.419~424 (2014)
- N. A. C. Cressie, "Statistics for Spatial Data" (1993), Jhon Wiley & Sons
- R. H. Myers, "Response Surface Methodology: Process and Product Optimization Using Designed Experiments" (1995), John Wiley & Sons
- Clive Loader, "Local Regression and Technologies" (1999), Springer
- G. Peano, "Sur nue Courbe qui Remplit Toute Aire Plane", Math. Ann, 36, pp.157~160 (1890)
- D. Hilbert, "Ueber die steitge Abbildung einer Linie auf ein Flaechenstueck", Math. Ann., 38, pp.459~460 (1891)