Enhanced two-point diagonal quadratic approximation methods for design optimization

被引:19
|
作者
Kim, Jong-Rip [3 ]
Choi, Dong-Hoon [1 ,2 ]
机构
[1] FRAMAX Co Ltd, Seoul 133791, South Korea
[2] Hanyang Univ, Ctr Innovat Design Optimizat Technol, Seoul 133791, South Korea
[3] Samsung Electromech Co Ltd, CAE Grp, Corp R&D Inst, Suwon, Gyunggi Do, South Korea
关键词
function approximation; two-point approximation; quadratic approximation; sequential approximate optimization;
D O I
10.1016/j.cma.2007.09.014
中图分类号
T [工业技术];
学科分类号
08 ;
摘要
Based on two-point diagonal quadratic approximation (referred to as TDQA), developed by Kim et al. [Min-Soo Kim, Jong-RipKim, Jae-Young Jeon, Dong-Hoon Choi, Efficient mechanical system optimization using two-point diagonal approximation in the non-linear intervening variable space, J. KSME 15 (2001) 1257-126], enhanced two-point approximation methods are proposed in this paper. The suggested methods reinforce TDQA with new quadratic correction terms using the concept of TANA-3. These methods overcome the disadvantage of TDQA when the derivatives at two design points have the same sign. In addition, the values and derivatives of the proposed approximation functions are completely equal to those of an original function at the two design points whether the derivatives at those points have the same sign or not. Several examples show the numerical performance and accuracy of the proposed methods compared to previous work, and optimization examples show that these methods can successfully reach the optimum via sequential approximation optimization. (c) 2007 Elsevier B.V. All rights reserved.
引用
收藏
页码:846 / 856
页数:11
相关论文
共 50 条
  • [1] A globally convergent sequential convex programming using an enhanced two-point diagonal quadratic approximation for structural optimization
    Seonho Park
    Seung Hyun Jeong
    Gil Ho Yoon
    Albert A. Groenwold
    Dong-Hoon Choi
    [J]. Structural and Multidisciplinary Optimization, 2014, 50 : 739 - 753
  • [2] A globally convergent sequential convex programming using an enhanced two-point diagonal quadratic approximation for structural optimization
    Park, Seonho
    Jeong, Seung Hyun
    Yoon, Gil Ho
    Groenwold, Albert A.
    Choi, Dong-Hoon
    [J]. STRUCTURAL AND MULTIDISCIPLINARY OPTIMIZATION, 2014, 50 (05) : 739 - 753
  • [3] Sequential Approximate Optimization by Dual Method Based on Two-Point Diagonal Quadratic Approximation
    Park, Seonho
    Jung, Sangjin
    Jeong, Seung-hyun
    Choi, Dong-Hoon
    [J]. TRANSACTIONS OF THE KOREAN SOCIETY OF MECHANICAL ENGINEERS A, 2011, 35 (03) : 259 - 266
  • [4] Efficient mechanical system optimization using two-point diagonal quadratic approximation in the nonlinear intervening variable space
    Min-Soo Kim
    Jong-Rip Kim
    Jae-Young Jeon
    Dong-Hoon Choi
    [J]. KSME International Journal, 2001, 15 : 1257 - 1267
  • [5] Enhanced Two-Point Exponential Approximation for Structural Optimization
    Chung, Tien-Tung
    Jiang, Qi-Hong
    Zhu, Jia-Rong
    [J]. ADVANCED DESIGN AND MANUFACTURING TECHNOLOGY III, PTS 1-4, 2013, 397-400 : 1021 - 1024
  • [6] Efficient mechanical system optimization using two-point diagonal quadratic approximation in the nonlinear intervening variable space
    Kim, MS
    Kim, JR
    Jeon, JY
    Choi, DH
    [J]. KSME INTERNATIONAL JOURNAL, 2001, 15 (09): : 1257 - 1267
  • [7] Effective two-point function approximation for design optimization
    Xu, SQ
    Grandhi, RV
    [J]. AIAA JOURNAL, 1998, 36 (12) : 2269 - 2275
  • [8] Two-point mid-range approximation enhanced recursive quadratic programming method
    Magazinovic, G
    [J]. STRUCTURAL AND MULTIDISCIPLINARY OPTIMIZATION, 2005, 29 (05) : 398 - 405
  • [9] Two-point mid-range approximation enhanced recursive quadratic programming method
    G. Magazinović
    [J]. Structural and Multidisciplinary Optimization, 2005, 29 : 398 - 405
  • [10] Numerical experience with two-point adaptive nonlinearity approximation for design optimization
    Magazinovic, G
    [J]. Design 2004: Proceedings of the 8th International Design Conference, Vols 1-3, 2004, : 1431 - 1434