Tool Wear Status Recognition Based on Wavelet Packet and 1D CNN
In order to monitor the non-linear wear changes of the tool during the cutting process of machine tools,a tool wear sta-tus recognition method based on wavelet packet decomposition and one-dimensional convolutional neural network(1D CNN)is proposed.The vibration data of machine tool spindle are collected as monitoring signals,and the wavelet packet after quantitative analysis of the signal-to-noise ratio is used for data preprocessing.Then the energy characteristics of each frequency band after the wavelet packet decomposition are selected as the input of 1D CNN to realizes effective identification of the tool wear status.Ex-perimental results show that the model can accurately prediction the tool wear status.Compared with BP network,energy spectrogram-Alexnet and Lstm network models,the recognition rate of tool wear status is the best,with an average accuracy rate of98.262%.
万方数据
维普
