人工神经网络在监测刀具切削状态中的应用

Application of Artificial Neural Network for the Monitoring of the State Cutting-Tools

基于人工神经网络（模式联想器）提出一种刀具磨损模型．训练时该网络把由多个传感器组成的输入矢量与由实测刀具磨损量组成的输出矢量联系起来，两矢量均作用于网络．实际运用时根据训练时已建立的“模型”，融合多个传感器的数据，给出刀具磨损的估计值．

This paper proposes a tool wear model based on Artificial Neural Network Which is used as Pattern Associator.As such,the network is trained to associate an input vector,consists of readings of several different sensors,with an output vector,consists of actual tool wear measurements,by a training process during which both vectors are presented to the network.During operation,the network is fed with the sensors'readings,and fuses this data and provides an estimate value for the wear according to the“model”established during the training.The estimation of tool wear obtained by this method is by far more accurate than other methods which use the same sensory inputs.