Engineering Design Based on Hammersley Sequences Sampling Method and SVR

Abstract:

Nowadays, computer-based engineering design is becoming widely used in the design of products. In the field of engineering, CAD model, FEM method and surrogate model are used to reduce the time and computational cost comparing with the traditional engineering design. But when facing with computationally expensive tasks, such design method mentioned above seems unable to deal with such tasks. In that case, surrogate model is gradually used and shows great potential in dealing with the computationally complex tasks. Although computing power and speed are rapidly growing, the use of the computer simulation analysis is limited in doing engineering design and some other analysis such as reliability analysis for complex product, so that it limits the use of metamodeling techniques. In that case, we use space-filling DOE sample method to support the construction of surrogate model. In this paper, we consider both Hammersley sequences and SVM as sampling method and surrogate model to construct the simulation design model, aiming at reduce computational costs. SVR achieves more accuracy and shows great potential in application in the design of complex and computationally expensive tasks.

• Periodical: Advanced Materials Research (Volume 544)
• Pages: 206-211
• DOI: https://doi.org/10.4028/www.scientific.net/AMR.544.206
• Online since: June 2012
• Authors: Li Ke, Hao Bo Qiu, Zhen Zhong Chen, Li Chi
• Keywords: Engineering Design, Hammersley Sequences, SVR

References

[1] Sacks, J., Welch, W. J., Mitchell, T. J., and Wynn, H. P., 1989, Design andAnalysis of Computer Experiments, Stat. Sci., 4.

[4] pp.409-435.

[2] Cristianni, N., and Shawe-Taylor, J., 2000, An Introduction to Support Vector Machines and Other Kernel-Based Learning Methods, Cambridge UniversityPress, Cambridge, UK.

DOI: 10.1017/cbo9780511801389

[3] Vapnik, V., and Lerner, A., 1963, Pattern Recognition Using Generalized Portrait Method, Autom., Remote Control (Engl. Transl. ), 24(6), pp.774-780.

[4] Vapnik, V., 1995, The Nature of Statistical Learning Theory, Springer, NewYork.

[5] Dumais, S. T., Platt, J., Heckerman, D., and Saharni, M., 1998, Inductive Learning Algorithms and Representations for Text Categorization, Proc. Of 7th Int. Conference on Information and Knowledge Management, Bethesda, MD, ACM, New York, pp.148-155.

DOI: 10.1145/288627.288651

[6] Fang H, Horstemeyer MF (2006), Global response approximation with radial basis functions, J Eng Optim 38(4): 407-424.

DOI: 10.1080/03052150500422294

[7] Martin JD, Simpson TW (2005), Use of kriging models to approximate deterministic computer models, AIAA J 43(4): 853-863.

DOI: 10.2514/1.8650

[8] Shao T, Krishnamurty S, 2008, A Clustering-Based Surrogate Model Updating Approach to Simulation-Based Engineering, Journal of Mechanical Design, Vol. 130 / 041101: 1-13.

DOI: 10.1115/1.2838329

[9] Clarke S. M, Griebsch J. H, Simpson T. W, (2005) Analysis of Support Vector Regression for Approximation of Complex Engineering Analyses, Journal of Mechanical Design, Vol. 127 pp: 1077-1087.

DOI: 10.1115/1.1897403

[10] Schölkopf, B., and Smola, A. J., 2002 Learning with Kernels: Support Vector Machines, Regularization, Optimization, and Beyond, MIT Press, Cambridge, MA.

DOI: 10.7551/mitpress/4175.001.0001