基于人工神经网络的变切削条件下钻头磨损监控

Research on Drill Wear Monitoring Under Varying Cutting Condition Based on Artificial Neural Network

基于人工神经网络建立变切削条件下的钻头磨损监控系统。以机床主轴和进给轴的电机功率（电流）信号为监控信号，并通过机床的速度向量识别机床的加工状态；通过对监控信号的提取和预处理，得到人工种经网络模型的输入（有效切谢功率和切削用量）；用3层BP网络对钻头的磨损量进行预报。

Drill wear monitoring system un-der varying cutting conditions based on artificialneural networks is established. The main motorpower and feed motor power of the machine toolare used as monitoring signals and the machinespeed vectors are used to recognize the workingstatus. This system gets the inputs of ANN (netcutting powers and machining conditions ) afterextracting and preprocessing the monitoring sig-nals. Three layers BP network is applied to mon-itor drill wear. Theoretical analysis and experi-mentaI results indicate that ANN model can beused in cutting tool wear monitoring under vary-ing cutting conditions and has high predictionprecision. The monitoring system possesses goodreliability and adaptability.