Micro-Milling Tool Wear Condition Monitoring Based on Semi-Supervised Parallel Gated CNN-LSTM
In the process of micro milling,the characteristics of high spindle speed and small tool size lead to serious tool wear.In order to achieve high-precision and high-efficiency tool wear condition monitoring,a monitoring method combining semi-supervised network with wavelet denoising and improved parallel ga-ted network is proposed.Firstly,the wavelet soft threshold function is used to remove noise and reduce the misleading degree of semi-supervised network for unlabeled data classification.Secondly,the semi-super-vised method uses labeled data for training to extract features and classify unlabeled data.Finally,the paral-lel gated convolutional neural network-long short-term memory network(CNN-LSTM)model is improved to extract global features and increase the expression ability of the model.The results show that the semi-su-pervised network after wavelet denoising can effectively increase the utilization rate of unlabeled data;The improved parallel gated CNN-LSTM model proposed by has a tool wear classification accuracy of 93.61%,which effectively improves the accuracy and efficiency of tool wear condition monitoring.
万方数据
维普
