Neural networks and response surface polynomials for design of vehicle joints

Abstract Typically, design of a complex system starts by setting targets for its performance characteristics. Then, design engineers cascade these targets to the components and design the components to meet these targets. It is important to have efficient tools that check if a set of performance targets for a component corresponds to a feasible design and determine the dimensions and mass of this design. This paper describes a method to develop tools that relate response parameters that describe the performance of a component to the physical design variables that specify its geometry. Neural networks and response surface polynomials are used to rapidly predict the performance characteristics of the components given the component dimensions. The method is demonstrated on design of an automotive joint. The paper compares neural networks and response surface polynomials and shows that they are almost equally accurate for the problem considered.

[1]  Anthony A. Giunta,et al.  Aircraft Multidisciplinary Design Optimization using Design of Experiments Theory and Response Surface Modeling Methods , 1997 .

[2]  Z. Peter Szewczyk.  A KNOWLEDGE-BASE D ADAPTIVE CLUSTERING TECHNIQUE FOR NEUROCOMPUTING MODELING OF STRUCTURES* , 1998 .

[3]  F ROSENBLATT,et al.  The perceptron: a probabilistic model for information storage and organization in the brain. , 1958, Psychological review.

[4]  Averill M. Law,et al.  Simulation Modeling and Analysis , 1982 .

[5]  D.R. Hush,et al.  Progress in supervised neural networks , 1993, IEEE Signal Processing Magazine.

[6]  C. L. Mallows.  Some comments on C_p , 1973 .

[7]  James L. Rogers,et al.  SIMULATING STRUCTURAL ANALYSIS WITH NEURAL NETWORK , 1994 .

[8]  W. Carpenter,et al.  A comparison of polynomial approximations and artificial neural nets as response surfaces , 1993 .

[9]  T. J. Mitchell,et al.  Bayesian Prediction of Deterministic Functions, with Applications to the Design and Analysis of Computer Experiments , 1991 .

[10]  Norman R. Draper,et al.  Applied regression analysis (2. ed.) , 1981, Wiley series in probability and mathematical statistics.

[11]  H. Akaike.  A new look at the statistical model identification , 1974 .

[12]  Asme,et al.  A collection of technical papers : the 32nd AIAA/ASME/ASCE/AHS/ASC Structures, Structural Dynamics, and Materials Conference, April 8-10, 1991 , 1991 .

[13]  Sidney Addelman,et al.  trans-Dimethanolbis(1,1,1-trifluoro-5,5-dimethylhexane-2,4-dionato)zinc(II) , 2008, Acta crystallographica. Section E, Structure reports online.

[14]  N. Draper,et al.  Applied Regression Analysis , 1966 .

[15]  Jasbir S. Arora,et al.  Introduction to Optimum Design , 1988 .

[16]  K. Schittkowski.  NLPQL: A fortran subroutine solving constrained nonlinear programming problems , 1986 .

[17]  B. Grossman,et al.  Variable-complexity response surface approximations for wing structural weight in HSCT design , 1996 .