Application and performance analysis of neural networks for decision support in conceptual design

This paper analyzes the use of feedforward multilayer perceptron neural networks in the support of the decision process during the conceptual design phase of engineering systems. A user friendly software tool is proposed in order to increase the quality of the designers' decisions by offering intuitive insights during the early design phase. The performance gain is evaluated through controlled experiments performed with non-experienced users. This paper proposes using the prediction capabilities of Artificial Neural Networks to recommend design solutions based on earlier successful designs. It implements a feedforward multilayer perceptron architecture with sigmoid activation functions, and uses the Levenberg-Marquardt training algorithm for weight matrix update. The network complexity was abstracted to ease the utilization by non-expert users. Different stopping conditions were developed including one where the training and testing error divergence is monitored to tackle the bias/variance dilemma. The advantages of this approach for decision support were measured through a set of six different case studies such as gaseous automobile emissions prediction based on institutional data, or the choice of a launcher for a specific space mission. The decision support tool generated quality performance gains between 13% and 88% for examples ranging from simple continuous single variable to complex discrete multivariate problems.

[1]  H. Robbins.  Some aspects of the sequential design of experiments , 1952 .

[2]  Dong Sik Jang,et al.  Approximate Estimation of the Product Life Cycle Cost Using Artificial Neural Networks in Conceptual Design , 2002 .

[3]  Flora Samuel,et al.  Designing in quality , 2010 .

[4]  H. Seifert,et al.  Rocket Propulsion Elements , 1963 .

[5]  Wei-Yin Loh,et al.  A Comparison of Prediction Accuracy, Complexity, and Training Time of Thirty-Three Old and New Classification Algorithms , 2000, Machine Learning.

[6]  Heaton T. Jeff,et al.  Introduction to Neural Networks with Java , 2005 .

[7]  Erman Aktürk,et al.  Artificial neural network-based model for predicting VO2max from a submaximal exercise test , 2011, Expert Syst. Appl..

[8]  Nancy Reid,et al.  On multiple comparisons among K samples subject to unequal patterns of censorship , 1977 .

[9]  H.A. Abyaneh,et al.  Interpolation of measured data based on neural network to model overcurrent relays in power systems , 2004, 2004 5th Asian Control Conference (IEEE Cat. No.04EX904).

[10]  A. Huxley,et al.  Hodgkin and the action potential 1935–1952 , 2002, The Journal of physiology.

[11]  N. Breslow.  A generalized Kruskal-Wallis test for comparing K samples subject to unequal patterns of censorship , 1970 .

[12]  Karl T. Ulrich,et al.  Product Design and Development , 1995 .

[13]  M. V. Rossum,et al.  Quantitative investigations of electrical nerve excitation treated as polarization , 2007, Biological Cybernetics.

[14]  David Enke,et al.  The use of data mining and neural networks for forecasting stock market returns , 2005, Expert Syst. Appl..

[15]  Kazuyuki Murase,et al.  Faster Training Using Fusion of Activation Functions for Feed Forward Neural Networks , 2009, Int. J. Neural Syst..

[16]  Alfredo Vellido,et al.  Neural networks in business: a survey of applications (1992–1998) , 1999 .

[17]  Michael A. Arbib,et al.  The handbook of brain theory and neural networks , 1995, A Bradford book.

[18]  J. J. Hopfield,et al.  “Neural” computation of decisions in optimization problems , 1985, Biological Cybernetics.

[19]  D. Marquardt.  An Algorithm for Least-Squares Estimation of Nonlinear Parameters , 1963 .

[20]  Willi-Hans Steeb.  Nonlinear Workbook: Chaos, Fractals, Cellular Automata, Neural Networks, Genetic Algorithms, Gene Expression , 2005 .

[21]  J Disney,et al.  Designing for quality , 2001 .

[22]  L. Lapicque.  Quantitative investigations of electrical nerve excitation treated as polarization. 1907. , 2007, Biological cybernetics.

[23]  Mukta Paliwal,et al.  Neural networks and statistical techniques: A review of applications , 2009, Expert Syst. Appl..

[24]  Özgür Kişi,et al.  Multi-layer perceptrons with Levenberg-Marquardt training algorithm for suspended sediment concentration prediction and estimation / Prévision et estimation de la concentration en matières en suspension avec des perceptrons multi-couches et l’algorithme d’apprentissage de Levenberg-Marquardt , 2004 .

[25]  Claire H. Jarvis,et al.  A Java-based intelligent advisor for selecting a context-appropriate spatial interpolation algorithm , 2004, Comput. Geosci..

[26]  O. Ks.  Multi-layer perceptrons with Levenberg-Marquardt training algorithm for suspended sediment concentration prediction and estimation , 2004 .

[27]  Elie Bienenstock,et al.  Neural Networks and the Bias/Variance Dilemma , 1992, Neural Computation.

[28]  F. G Parsons,et al.  A Kolmogorov - Smirnov goodness-of-fit test for the two-parameter weibull distribution when the parameters are estimated from the data , 1982 .

[29]  Willi-Hans Steeb,et al.  The Nonlinear Workbook , 2005 .