摘要
通过分析总结影响刀具寿命的主要影响因素,建立铣削刀具加工参数与刀具寿命的径向基神经网络模型。训练模型使用了10组样本数据,以刀具直径、铣削速度、铣削宽度、铣削深度、进给量、刀具齿数作为网络输入参数,采用十折交叉验证方法对所构建模型进行验证,能够对刀具寿命进行较为准确的预测。与传统BP神经网络模型比较发现,径向基神经网络具有更好的预测精度和稳定性,是预测刀具寿命的一条有效途径。
By analyzing and summarizing the main influencing factors of tool life, a radial basis neural network model for machining parameters and life of milling tools is established. The training model uses 10 sets of sample data. The tool diameter, milling speed, milling width, milling depth, feed and tool teeth number are used as network input parameters. The ten-fold cross validation method is used to validate the model and the tool life can be accurately predicted. Compared with the traditional BP neural network model, the RBF neural network has better prediction accuracy and stability, and is an effective approach to tool life prediction.
引文
[1]王杰,李方信,肖素梅.机械制造工程学[M].北京:北京邮电大学出版社,2015.
[2]丁怡,何卫平,张维,王伟,李亚杰,董荣.基于BP神经网络的刀具寿命预测模型[J].航空制造技术,2010(8):93-96.
[3]徐玲,杣丹,王时龙,聂建林.基于进化神经网络的刀具寿命预测[J].计算机集成制造系统,2008(1):167-171,182.
[4]张德丰.MATLAB神经网络编程[M].北京:化学工业出版社,2011.
[5]黄媛,孙树栋,李兢尧.基于ACO-BP神经网络的刀具寿命预测[J].机械科学与技术,2009,28(11):1517-1521.
[6]Pan M,Tang W,Xing Y,et al.The Layout Measurement Points Prediction and Flatness Calculation for the Antenna Plate After Welding Assembly Based on the RBF Neural Network Model[C].ASME 2016 International Design Engineering Technical Conferences and Computers and Information in Engineering Conference,2016:V004T05A001.
[7]张永志,董俊慧,张艳飞.基于径向基神经网络焊接接头力学性能预测[J].焊接学报,2008(7):81-84,117.
[8]张梅军,王闯,陈灏.IMF能量和RBF神经网络相结合在滚动轴承故障诊断中的应用研究[J].机械,2012,39(6):63-66,70.
[9]Noemia G D M D M,Ferraz A Q.Life Prediction of Cutting Tool by the Workpiece Cutting Condition[J].Advanced Materials Research,2011(223):554-563.
[10]MATLAB中文论坛.MATLAB神经网络30个案例分析[M].北京:北京航空航天大学出版社,2010.