Tuning of Shape Parameters for Radial Basis Functions in Nonlinear Meta-Modeling

Abstract:

Meta-modeling is a frequently used tool for analysis of systems' behavior. An original model of the system is often complex and its evaluation is expensive and time-consuming. Therefore it is desirable to execute the original model as few times as possible. A special case is when many evaluation of the model with different input parameters are necessary. Proposed solution is a use of the meta-model, in our case Radial Basis Function Network (RBFN) tool is presented. Here, the output of the meta-model is constructed as a linear combination of the radial basis functions. For good approximation a shape parameter of the radial basis functions has to be set properly. This paper describes a tuning of the shape parameters for several benchmark examples.

• Periodical: Applied Mechanics and Materials (Volume 825)
• Pages: 149-152
• DOI: https://doi.org/10.4028/www.scientific.net/AMM.825.149
• Online since: February 2016
• Authors: Eva Myšáková*, Matěj Lepš
• Keywords: Meta-Modeling, Radial Basis Function Network, Shape Parameter, Stochastic Optimization

References

[1] A. Forrester, A. Sóbester, A. Keane, Engineering Design via Surrogate Modelling: A Practical Guide, John Wiley & Sons Ltd, ISBN 978-0-470-06068-1, (2008).

DOI: 10.1002/9780470770801

[2] D. -C. Montgomery, Design and Analysis of Experiments, Wiley, 8 edition, ISBN: 978- 1118146927, (2012).

[3] A. -B. Ryberg, Metamodel-Based Design Optimization: A Multidisciplinary Approach for Automotive Structures, Licentiate thesis, Linköping UniversityLinköping University, Solid Mechanics, The Institute of Technology, (2013).

[4] R. Jin, W. Chen, T. Simpson, Comparative studies of metamodelling techniques under multiple modelling criteria. Structural and Multidisciplinary Optimization, 23 (1) 1-13, (2001).

DOI: 10.1007/s00158-001-0160-4

[5] J. Andre, P. Siarry, T. Dognon, An improvement of the standard genetic algorithm fighting premature convergence in continuous optimization. Advances in Engineering Software, 32 (1) 49- 60, (2000).

DOI: 10.1016/s0965-9978(00)00070-3

[6] N. Dyn, D. Levin, S. Rippa, Numerical Procedures for Surface Fitting of Scattered Data by Radial Basis Functions, SIAM Journal of Scientific and Statistical Computing, 7 (2) 639-659, (1986).

DOI: 10.1137/0907043

[7] A. Kučerová, Identification of nonlinear mechanical model parameters based on softcomputing methods, PhD. thesis, Ecole Normale Supérieure de Cachan, Laboratoire de Mécanique et Technologie, (2007).

[8] H. Nakayama, K. Inoue, Y. Yoshimori, Approximate optimization using computational intelligence and its application to reinforcement of cable-stayed bridges. In European Congress on Computational Methods in Applied Sciences and Engineering (ECCOMAS 2004), Jyväskylä, (2004).

[9] R. R. Picard, R. D. Cook, Cross-Validation of Regression Models, Journal of the American Statistical Association, 79 (387) 575-583, (1984).

DOI: 10.1080/01621459.1984.10478083

[10] S. H. Lee, B. M. Kwak, Reliability based design optimization using response surface augmented moment method, (2005).