Prediction of the Cutting Depth of Abrasive Suspension Jet Using a BP Artificial Neural Network

Abstract:

Abrasive suspension jet is a new embranchment of abrasive jet. In the cutting process of this jet, the suspension concentration is constant, so the cutting quality is more stable. In this paper, a prediction model based on a back-propagation (BP) artificial neural network is presented for predicting the cutting depth generated by abrasive suspension jet. In the application of the BP neural network, the mean error of the output in the model training is 0.01, the relatively discrepancy is below 8.70%. The modeling method based on the BP neural network is much more convenient and exact compared with traditional methods, and can always achieve a much better prediction effect. It is verified with experiments to be reasonable and feasible, and it is the better foundation for the future study of abrasive suspension jet.

• Periodical: Advanced Materials Research (Volume 500)
• Pages: 249-252
• DOI: https://doi.org/10.4028/www.scientific.net/AMR.500.249
• Online since: April 2012
• Authors: Xiao Jian Liu, Qian Qian Fan, Yan Xia Feng
• Keywords: Abrasive Suspension Jet, Cutting Depth, Neural Network (NN)

References

[1] J.M. Cabrera, A. Alomar, J.J. Jonas, J.M. Prado, Metall. Trans. A 28 (2007) 2233±2244.

[2] Xiao-jian Liu Tao Yu , Study on the Theory of A New Kind of Abrasive Jet Cutting Technology and the Test Analyses, Machinery , 2005. 2.

[3] Refenes A. N. Stock performance modeling using neural networks : A Comparative study with regression models, Neural Networks, 2004, 17 (2).

DOI: 10.1016/0893-6080(94)90030-2

[4] L. Monostori and D. Barschdorff, Artificial neural networks in intelligent manufacturing, Robotics and Computer Integrated Manufacturing 9, 412-436 (2009).

[5] Taylor. J.G. (1998) Neural Networks : an overview in L.J. Landau, &J.F. Taylor. Concepts for neural networks – a Survey London : Springer.