基于改进GRU神经网络的刀具磨损状态预测

打开文本图片集

关键词：刀具磨损;深度学习;卷积神经网络（CNN）;门控循环单元（GRU）;注意力模块;状态预测中图分类号：TP391.4;TP183文献标识码：Adoi：10.37188/OPE.20253323.3765 CSTR：32169.14.OPE.20253323.3765

Abstract：Toaddress the challenges ofcumbersome signal feature extraction in traditional intelligent monitoring methods and the adverse efects of tool wear on workpiece quality and production eficiency during millig，a tool wear state prediction method based on an improved GRU neural network （BiGRU1DCNN-CBAM） is proposed. Using statistical methods，time-domain analysis，frequency-domain analysis，and wavelet transform， 24 feature parameters of the tool signals-such as mean，kurtosis，and power spectraldensity-are extracted，and the resulting multimodal time-series data are converted into time-series images of tool features.A convolutional neural network（CNN） is then introduced to mine deep features of the signal data，and a convolutional block atention module （CBAM） is integrated to enhance the capability of the model to capture feature maps of vibration and cuting force signals. After flattning and concatenating the feature layers，the fused features are fed into a bidirectional GRU （BiGRU） to capture longterm dependencies，and the tool wear amount is predicted through a fully connected layer，thereby enabling remaining useful life prediction of the tool wear state during CNC machining. Experimental results on the PHM2OlO dataset show that the RMSE and MAE of the proposed model are 2.17μm and 1. 29 μm ， respectively. Compared with Bayesian-MCMC-Prognostics， SBiLSTM，RIME-CNN-SVM，MobileNetV3， TDConvLSTM， ISABO-IBiLSTM，IWOA-IECA-BiLSTM，and LSTM-CNN-CBAM models，the prediction accuracy in terms of RMSE and MAE is improved by more than 40.5% and （204号 52.1% ，respectively，while the time consumption is reduced by at least 2.8% relative to similar models. These results demonstrate that the proposed model can efectively characterize tool wear，reduce prediction errors，and achieve superior prediction performance.

Key words： tool wear;deep learning；Convolutional Neural Network （CNN）；Gate Recurrent Unit （GRU）;attention module；state prediction

1引言

随着现代工业自动化的发展，预测与健康管理技术在制造系统中得到广泛应用。（剩余25616字）