Response surface approximations for discipline coordination in multidisciplinary design optimization

Sellar, Richard S.; Stelmack, Marc A.; Batill, Stephen M.; Renaud, John E.

doi:10.2514/6.1996-1383

Cited by 27 publications

(9 citation statements)

References 5 publications

Supporting

Mentioning

Contrasting

Order By: Relevance

“…of unknown coeffi- » (P),(P) (7) Note that there is a difference between the p th vector of independent variables, x( p ), and the p th vector of independent variables mapped into the form of the polynomial model, x( p ).…”

Section: Polynomial Approximation Modelsmentioning

confidence: 99%

See 1 more Smart Citation

A comparison of approximation modeling techniques - Polynomial versus interpolating models

Anthony

Layne

1998

7th AIAA/USAF/NASA/ISSMO Symposium on Multidisciplinary Analysis and Optimization

295

145

View full text Add to dashboard Cite

NomenclatureTwo methods of creating approximation models are compared through the calculation of the modeling accuracy on test problems involving one, five, and ten independent variables. Here, the test problems are representative of the modeling challenges typically encountered in realistic engineering optimization problems. The first approximation model is a quadratic polynomial created using the method of least squares. This type of polynomial model has seen considerable use in recent engineering optimization studies due to its computational simplicity and ease of use. However, quadratic polynomial models may be of limited accuracy when the response data to be modeled have multiple local extrema. The second approximation model employs an interpolation scheme known as kriging developed in the fields of spatial statistics and geostatistics. This class of interpolating model has the flexibility to model response data with multiple local extrema. However, this flexibility is obtained at an increase in computational expense and a decrease in ease of use. The intent of this study is to provide an initial exploration of the accuracy and modeling capabilities of these two approximation methods.

show abstract

Section: Polynomial Approximation Modelsmentioning

confidence: 99%

“…In addition, the method of least squares used in creating the models is relatively computationally inexpensive and straightforward. Evidence of the popularity of polynomial modeling is given by the wealth of recent reports both by the authors of this document [2] and by others as well (cf., [3,4,5,6,7,8,9]). …”

Section: Introductionmentioning

confidence: 98%

A comparison of approximation modeling techniques - Polynomial versus interpolating models

Anthony

Layne

1998

7th AIAA/USAF/NASA/ISSMO Symposium on Multidisciplinary Analysis and Optimization

295

145

View full text Add to dashboard Cite

show abstract

“…The objective is to determine the least gross take-off weight design within the bounded design space, subject to two performance constraints. The MDO formulation of the aircraft concept sizing problem is posed as Minimize: W total W empty • W f uel (16) •W paylaod Subject to: g 1 Range Range required 1 0 0 0 It is observed that the RS-CSSO method takes far fewer iterations to converge to the same optimum as obtained by LANCELOT. The solution reached by both algorithms is recognized as the global optimum of the problem.…”

Section: Aircraft Concept Sizing Problemmentioning

confidence: 99%

A comparative study of trust region managed approximate optimization

Lin

Renaud

2001

19th AIAA Applied Aerodynamics Conference

View full text Add to dashboard Cite

This research investigates two competing strategies for managing the interaction between the optimization and the fidelity of the approximation models. Effective management ensures that the process converges to a solution of the original design problem. Two trust region managed approximate optimization approaches are studied in this research: a response surface based concurrent subspace optimization (RS-CSSO) strategy and a commercially available software package LANCELOT. A detailed performance comparison is conducted to evaluate how the two methods perform on different classes of optimization problems. A series of optimization problems ranging from simple analytic codes to multidisciplinary coupled engineering test problems are used. Both methods sequentially optimize approximate models of the augmented Lagrangian function subject to a trust region constraint. The RS-CSSO method builds a cumulative response surface approximation constructed from variable fidelity design data generated during CSSO's, while LANCELOT uses a Broyden-Fletcher-Goldfarb-£ Product Design Engineer † Associate Professor, Associate Fellow AIAA Copyright ©2001 by John E. Renaud. Published by the American Institute of Aeronautics and Astronautics, Inc. with permission. Shanno (BFGS) quadratic approximation. The performances of both optimizers are compared using the optima obtained and the number of optimization iterations as metrics. Results indicate that the response surface based CSSO strategy exhibits superior performance in application to multidisciplinary design optimization test problems.

show abstract

“…Additionally, approximation models and sensitivity analysis are implemented and applied to predict missing data in the database. [5][6][7] Analysis methods are developed for aerodynamics, propulsion and performance. [8][9][10] These analysis codes are implemented in a spreadsheet using the well-known Visual Basic for applications (VBA) programming language for simple and efficient calculation.…”

Section: Process and Methodologymentioning

confidence: 99%

Aircraft Design Optimization Process using a Decision Model by Database

Naing

Nguyen

Lee

et al. 2015

Trans. Japan Soc. Aero. S Sci.

View full text Add to dashboard Cite

In this study, an efficient methodology capable of systematically constructing an aircraft design database is developed and its application is discussed. The database focuses on fighter and attack aircraft because their design is a particular challenge compared with the design of other types of aircraft. For small conventional aircraft, historical information accumulated over many years on countless previous designs is freely available to designers. Such data for fighter aircraft design are, however, difficult to obtain or are insufficient. The database developed can contribute to the design of fighter and attack aircraft on a conceptual level. In addition, decision-making models are developed by means of data mining as well as approximation models based on the developed database to facilitate initial aircraft configuration decisions. The efficiency and simplicity of the proposed method are described in a case study, which also describes a sample process of conceptual design optimization using initial design information extracted from the developed database.

show abstract

Response surface approximations for discipline coordination in multidisciplinary design optimization

Cited by 27 publications

References 5 publications

A comparison of approximation modeling techniques - Polynomial versus interpolating models

A comparison of approximation modeling techniques - Polynomial versus interpolating models

A comparative study of trust region managed approximate optimization

Aircraft Design Optimization Process using a Decision Model by Database

Contact Info

Product

Resources

About