Real-time monitoring method for surface roughness of γ-TiAl alloy based on deep learning of time-frequency diagram

Abstract γ-TiAl alloy is a typically difficult material to machine, and machining defects such as grain pull-out and material spalling are common during machining, resulting in component scrap. As a result, it is critical to investigate the real-time monitoring method of surface roughness during milling. A new model for predicting surface roughness is proposed in this article. Based on the idea of deep learning, the prediction problem of surface roughness is turned into a classification problem. The features of the force signal are extracted using a continuous wavelet transform, and the one-dimensional signal is converted into a two-dimensional time-frequency diagram to obtain more information. The transfer learning mechanism is introduced to make the model run faster and improve prediction accuracy. The overfitting problem is solved and the model's generalization ability is improved through batch normalization and dropout layer, and the prediction accuracy is increased by 3%. The results show that the method has high recognition accuracy, with a maximum classification accuracy of 98% and an average accuracy of 96.12%, a 6% improvement over the traditional model, allowing it to accurately realize real-time monitoring of surface quality in milling processing.