Optimization Methods of High-Speed Machining Processing Based on Fuzzy Neural Network

Abstract:

A model of five layers of fuzzy neural network which is appropriate for high-speed machining (HSM) processing optimization is proposed to fully exert the advantages of HSM. This model built a mapping relationship among the credibility of three-optimization objective "efficiency priority", "quality priority" and "precision priority”, material removal rate, cutter rigidity, machine-tool precision, and the smooth degree and fitting error of tool path. Gradient descent method is used to conduct training of network on the basis of orthogonal test. The constraints are also given in the HSM processing optimization objective. A prototype system of HSM processing optimization is developed based on fuzzy neural network. The result of test shows that not only the systems runs reliably and steady, but also the processing scheme optimized is economic and reasonable.

• Periodical: Key Engineering Materials (Volume 464)
• Pages: 5-10
• DOI: https://doi.org/10.4028/www.scientific.net/KEM.464.5
• Online since: January 2011
• Authors: Quan Ping Sun
• Keywords: Fuzzy Neural Network (FNN), High Speed Machining (HSM), Optimization, Processing

References

[1] Zhai Yushan. Dissertation for ME from Hunan University. (2007), p.46.

[2] Savas Vedat, Ozay Cetin. The International Journal of Advanced manufacturing Technology. Vol. 37(2008), pp.335-340.

[3] Liu Changqing. Dissertation for PhD from Harbin Institute of Technology. (2007), p.61.

[4] Li J.G., Zhao H., Yao Y.X., Liu C.Q. International Journal of Advanced manufacturing Technology. Vol. 36(2008), p.908.

[5] A hmad Nafis, Wakabayashi Ryo, Tomohisa Tanaka et al . In: The 11th International Conference on Precision Engineering. (2007), p.347.

[6] Yang S.H., Natarajan. International of Journal Advanced manufacturing Technology. Vol. 49(2010), p.773.

[7] Wang Xiankui. Mechanical Processing Handbook. Beijing: Mechanical Industry Press. (2006), p.30.