A neural network approach for grinding processes: Modelling and optimization

Abstract The objective of this study is to show how back propagation (BP) neural networks can be used to model and optimize grinding processes, using creep feed grinding of alumina with diamond wheels as an example. First, a generalized back propagation neural network with two-hidden layers is used to establish the process model. Then the back propagation algorithm with Boltzmann factor is used to find the global optimal settings for the grinding process. From the simulation results obtained, it is found that the implemented neural network approach yields a more accurate process model than the regression method. It is also shown that, unlike the conventional back propagation network, proper use of the Boltzmann factor with BP can effectively avoid local minima and generate the global optimal solution.

[1]  Geoffrey E. Hinton,et al.  A Learning Algorithm for Boltzmann Machines , 1985, Cogn. Sci..

[2]  James L. McClelland,et al.  Parallel distributed processing: explorations in the microstructure of cognition, vol. 1: foundations , 1986 .

[3]  J. Hopfield,et al.  Computing with neural circuits: a model. , 1986, Science.

[4]  Geoffrey E. Hinton,et al.  Learning internal representations by error propagation , 1986 .

[5]  George Chryssolouris,et al.  A Comparison of Statistical and AI Approaches to the Selection of Process Parameters in Intelligent Machining , 1990 .

[6]  Stephen Grossberg,et al.  A massively parallel architecture for a self-organizing neural pattern recognition machine , 1988, Comput. Vis. Graph. Image Process..

[7]  Stephen Grossberg,et al.  Art 2: Self-Organization Of Stable Category Recognition Codes For Analog Input Patterns , 1988, Other Conferences.

[8]  J J Hopfield,et al.  Neural networks and physical systems with emergent collective computational abilities. , 1982, Proceedings of the National Academy of Sciences of the United States of America.

[9]  Jacques de Villiers,et al.  Backpropagation neural nets with one and two hidden layers , 1993, IEEE Trans. Neural Networks.

[10]  Zoran Obradovic,et al.  Small Depth Polynomial Size Neural Networks , 1990, Neural Computation.

[11]  Richard P. Lippmann,et al.  An introduction to computing with neural nets , 1987 .

[12]  T. Warren Liao,et al.  An evaluation of ART1 neural models for GT part family and machine cell forming , 1993 .

[13]  S.S. Rangwala,et al.  Learning and optimization of machining operations using computing abilities of neural networks , 1989, IEEE Trans. Syst. Man Cybern..

[14]  Teuvo Kohonen,et al.  Self-Organization and Associative Memory , 1988 .

[15]  László Monostori,et al.  Artificial neural networks in intelligent manufacturing , 1992 .

[16]  David Haussler,et al.  What Size Net Gives Valid Generalization? , 1989, Neural Computation.

[17]  Ming-Kuen Chen,et al.  Neural network modelling and multiobjective optimization of creep feed grinding of superalloys , 1992 .