基于模糊控制和BP神经网络的高速铣削恒力控制研究-BP.PDF

International Journal of Control and Automation Vol.7, No.5 (2014), pp.143-152 /10.14257/ijca.206 High Speed Constant Force Milling Based on Fuzzy Controller and BP Neural Network Yunlin Huang1 and Juntang Yuan1 1School of Mechanical Engineering, Nanjing University of Science and Technology Nanjing 210094, China mashuangyunlin@163.com, mc106@ L. Abstract A . In order to meet the high cutting rate and high surface quality requirement, put forward the y G fuzzy control, on-line adjusting the cutting feed rate method to realize a constant force in l E CNC high speed milling. The input variables of a fuzzy controller are the difference between n L the reference forces and the actual cutting forces and the change rate of the deviation, after O fuzzification, fuzzy inference and defuzzification, the output variable is the change of per L I tooth feed. The control strategy of on-line adjusting fuzzy rules and on-line self-adaptiving the n s output scaling factor is used in a fuzzy controller. The CNC high speed milling process model o i is established by 3-3-5-1