Deep learning-based in-situ monitoring of cutting forces in ultra-precision diamond turning

In ultra-precision diamond turning (UPDT), the effective monitoring of cutting force is promising to guarantee the desired surface quality. However, most of the monitoring methods require pre-processing of the original signal and it would induce data loss. This issue is especially serious during in-situ monitoring of cutting force signals with dynamic and nonlinear characteristics. Motivated by this, an in-situ cutting force monitoring method is proposed for UPDT based on deep learning. Firstly, a theoretical-fully connected neural network (TFCNN) model is established for simulating the cutting forces by combining a theoretical model and a fully connected neural network. Secondly, an in-situ monitoring model is proposed by using an autoencoder with the input signals of the simulated cutting forces according to the TFCNN model. Finally, to verify the effectiveness of the proposed model, a series of UPDT experiments are conducted and the cutting forces and the machined surfaces are measured correspondingly. The results reveal that the in-situ monitoring model has excellent performance in monitoring the anomalous data of the dynamic and nonlinear cutting forces. Through analysing the relationship among the machined surface, the cutting forces, and the relative errors, the overall accuracy of the proposed in-situ monitoring model can achieve 92.59%, which provides an effective and accurate method for in-situ monitoring of cutting forces in UPDT.