Comparison of optimization by response surface methodology with neurofuzzy methods

We compare two approaches where empirical models are used to augment computer simulations to facilitate rapid device optimization. We apply the response surface model (RSM) methodology and neurofuzzy techniques to the problem of modeling simulations of the average flux density in the air gap of a loudspeaker. Both these techniques have significant advantages over more traditional methods of optimizing computer simulation experiments. We show that these techniques have different advantages and disadvantages depending on the problem being modeled. In particular, the use of domain knowledge is shown to give robust and reliably predictive RSM's. Neurofuzzy techniques are shown to be particularly suited to problems where little is known about the problem.

[1]  Annette J. Dobson,et al.  An introduction to generalized linear models , 1991 .

[2]  David E. Goldberg,et al.  Genetic Algorithms in Search Optimization and Machine Learning , 1988 .

[3]  Robert Spence,et al.  Modelling with gradient information , 1999 .

[4]  J. A. Nelder,et al.  Inverse Polynomials, a Useful Group of Multi-Factor Response Functions , 1966 .

[5]  Chuen-Tsai Sun,et al.  Neuro-fuzzy modeling and control , 1995, Proc. IEEE.

[6]  Martin Brown,et al.  High dimensional neurofuzzy systems: overcoming the curse of dimensionality , 1995, Proceedings of 1995 IEEE International Conference on Fuzzy Systems..

[7]  Hua Su,et al.  Informative experimental design for electronic circuits , 1998 .

[8]  T Watson Layne,et al.  Multidisciplinary Optimization of a Supersonic Transport Using Design of Experiments Theory and Response Surface Modeling , 1997 .

[9]  Christian Magele,et al.  Approximation of the objective function: multiquadrics versus neural networks , 1999 .

[10]  R. Fletcher.  Practical Methods of Optimization , 1988 .

[11]  Bernhard Brandstätter,et al.  Utilizing feedforward neural networks for acceleration of global optimization procedures [SMES problems] , 1998 .

[12]  Hua Su,et al.  Applying response surface methodology in the design and optimization of electromagnetic devices , 1997 .

[13]  Stephen L. Chiu,et al.  Fuzzy Model Identification Based on Cluster Estimation , 1994, J. Intell. Fuzzy Syst..

[14]  Robert Spence,et al.  Response surface models of electromagnetic devices and their application to design , 1999 .

[15]  P. McCullagh,et al.  Generalized Linear Models , 1972, Predictive Analytics.

[16]  Maurizio Repetto,et al.  A combined approach for the stochastic optimisation of multiminima problems using adaptive fuzzy sets and radial basis functions , 1998 .

[17]  A. E. Ainsley,et al.  Genstat 5 Reference Manual , 1987 .

[18]  W. Edwards Deming,et al.  Out of the Crisis , 1982 .

[19]  Eva Riccomagno,et al.  Experimental Design and Observation for Large Systems , 1996, Journal of the Royal Statistical Society: Series B (Methodological).

[20]  Piergiorgio Alotto,et al.  A "Design of Experiment" Approach to Enhance the "Generalized Response Surface" Method in the Optimization of Multiminima Problems , 1997 .

[21]  Leland M. Nicolai.  Fundamentals of Aircraft Design , 1984 .

[22]  John H. Holland,et al.  Adaptation in Natural and Artificial Systems: An Introductory Analysis with Applications to Biology, Control, and Artificial Intelligence , 1992 .

[23]  Piergiorgio Alotto,et al.  A "design of experiment" and statistical approach to enhance the "generalised response surface" method in the optimisation of multiminima problems , 1997 .

[24]  M. Repetto,et al.  Multiobjective optimization in magnetostatics: a proposal for benchmark problems , 1996 .

[25]  Douglas O. Stanley,et al.  Approximation model building and multidisciplinary design optimization using response surface methods , 1996 .

[26]  P. McCullagh,et al.  Generalized Linear Models , 1992 .

[27]  Anthony J. Walton,et al.  The integration of simulation and response surface methodology for the optimization of IC processes , 1994 .

[28]  Z. Malik,et al.  Modified inverse polynomial as a response surface model for magnetic flux density , 1998 .

[29]  Repetto,et al.  Response Surface method in magnetic optimization , 1993 .

[30]  M. Matsunami,et al.  An optimization method based on radial basis function , 1997 .

[31]  George E. P. Box,et al.  Empirical Model‐Building and Response Surfaces , 1988 .

[32]  Robert Spence,et al.  Application of generalized linear models to the design improvement of an engineering artefact , 1996 .

[33]  K. Rashid,et al.  Optimisation of electromagnetic devices using computational intelligence techniques , 1999, IEEE International Magnetics Conference.

[34]  Robert Spence,et al.  Generalized linear models for empirical performance modeling in circuit design , 1994, Proceedings of APCCAS'94 - 1994 Asia Pacific Conference on Circuits and Systems.

[35]  Robert Spence,et al.  The facilitation of insight for analog design , 1999 .

[36]  Anthony J. Walton,et al.  Wafer mapping using DOE and RSM techniques , 1995, Proceedings International Conference on Microelectronic Test Structures.

[37]  David A. Lowther,et al.  Response surface modelling of magnetic device performance using function value and gradient , 1998 .

[38]  H. Wynn,et al.  Generic circuit response interpolation for robust design , 1995 .

[39]  M. Repetto,et al.  A multiquadrics-based algorithm for the acceleration of simulated annealing optimization procedures , 1996 .