International Journal of Computational Intelligence Research

Open Access Open Access  Restricted Access Subscription Access
Open Access Open Access Open Access  Restricted Access Restricted Access Subscription Access

Optimal Selection of Artificial Neural Network Parameters with Design of Experiments: An Application in Modeling Milling of Carbon Fiber Reinforced Composites


Affiliations
1 Mechanical Engineering Technology, Metropolitan State University of Denver Denver, CO, United States
2 Department of Industrial and Manufacturing Engineering, Wichita State University, Wichita, KS, United States
     

   Subscribe/Renew Journal


Artificial neural networks (ANN) provide superior solutions to a variety of modeling and optimization problems in manufacturing. However, determining suitable network parameters such as number of neurons, number of hidden layers and learning rate of an ANN still remains a difficult task. This paper presents the application of the analysis of variance (ANOVA) method for the optimization of ANN model and identifying the optimal ANN parameters. A case study of a modeling resultant cutting force in milling process is used to demonstrate implementation of the approach. This paper also presents an application of model to predict machining forces of carbon fiber reinforced polymer (CFRP) composites when data are highly dimensional and nonlinear. Statistical techniques were utilized to evaluate performance in terms of the ischolar_main mean squared error (RMSE) of predicted forces. The Analysis of Variance (ANOVA) indicated that the number of hidden layers has no significant effect on the prediction error. The effect of the transfer function was shown to have a dominating effect on the RMSE. In addition, the learning rule cannot be selected without reference to the transfer function used. The results indicated that ANN training is sensitive to changes in the transfer function and learning rule used.
Subscription Login to verify subscription
User
Notifications
Font Size


  • Kalla, D., Sheikh-Ahmad, J. & Twomey, J., (2010). “Prediction of Cutting Forces in Helical End Milling of Fiber Reinforced Polymers”, International Journal of Machine Tools & Manufacture, 50 (10), 882-891.

  • Sheikh-Ahmad J, Twomey J, Kalla D, Lodhia P (2007) Multiple regression and committee neural network force prediction models in milling FRP. Machining Science and Technology 11(3), 391-412.

  • Kalla, D., Nahusha, K. and Dhanasekaran, PS (2011). “Optimization of Cutting Forces in End Milling of CFRP composites using Response Surfaces and Desirability Function Approach”, SAMPE, LongBeach, CA, May 23-26.

  • Laosiritaworn, W. and Chotchaithanakorn, N (2009). "Artificial Neural Networks Parameters Optimization with Design of Experiments: An Application in Ferromagnetic Materials Modeling", Chiang Mai J. Sci. 36(1), pp.83-91.

  • Kim Y.S., and Yum B.J., Robust Design of Multilayer Feed Forward Neural Networks: An Experimental Approach, Appl. Artif. Intell., 2004; 17: 249-263.

  • Yang S.M., and Lee G.S., Neural Network Design by using Taguchi Method, J. Dyn. Syst. Meas. Contr. 1999; 121: 560-563.

  • Khaw J.F.C., Lim B.S., and Lim L.E.N., Optimal Design of Neural Networks using the Taguchi Method, Neuralcomputing, 1995; 7: 225-245.

  • Neural Works Professional II/Plus software, Neuralware, Pennsylvania.

  • Design Expert software, Version 8, Stat-Ease, Inc., Minneapolis.


Abstract Views: 320

PDF Views: 0




  • Optimal Selection of Artificial Neural Network Parameters with Design of Experiments: An Application in Modeling Milling of Carbon Fiber Reinforced Composites

Abstract Views: 320  |  PDF Views: 0

Authors

Devi K. Kalla
Mechanical Engineering Technology, Metropolitan State University of Denver Denver, CO, United States
Aaron Brown
Mechanical Engineering Technology, Metropolitan State University of Denver Denver, CO, United States
Nahusha Kumar
Department of Industrial and Manufacturing Engineering, Wichita State University, Wichita, KS, United States

Abstract


Artificial neural networks (ANN) provide superior solutions to a variety of modeling and optimization problems in manufacturing. However, determining suitable network parameters such as number of neurons, number of hidden layers and learning rate of an ANN still remains a difficult task. This paper presents the application of the analysis of variance (ANOVA) method for the optimization of ANN model and identifying the optimal ANN parameters. A case study of a modeling resultant cutting force in milling process is used to demonstrate implementation of the approach. This paper also presents an application of model to predict machining forces of carbon fiber reinforced polymer (CFRP) composites when data are highly dimensional and nonlinear. Statistical techniques were utilized to evaluate performance in terms of the ischolar_main mean squared error (RMSE) of predicted forces. The Analysis of Variance (ANOVA) indicated that the number of hidden layers has no significant effect on the prediction error. The effect of the transfer function was shown to have a dominating effect on the RMSE. In addition, the learning rule cannot be selected without reference to the transfer function used. The results indicated that ANN training is sensitive to changes in the transfer function and learning rule used.

References