A confocal axial monitoring method for femtosecond laser processing parameters based on deep learning

Abstract

The invention discloses a femtosecond laser processing parameter confocal axial monitoring method based on deep learning, which belongs to the technical field of laser precision processing detection, including: S1, constructing a laser processing parameter confocal axial monitoring network based on a convolutional neural network; S2. Use the training samples to train the confocal axial monitoring network of laser processing parameters; S3. Input the surface topography and surface real-time axial position of the workpiece, the real-time position of the precision workbench, and femtosecond laser spatio-temporal parameters into the training to complete The confocal axial monitoring network of the laser processing parameters is used to obtain the real-time femtosecond laser processing accuracy and shape change monitoring results of the workpiece. The invention combines the confocal axial monitoring system with the convolutional neural network to train the deep learning model in the case of large data, so that it can quickly and synchronously accurately focus and align during the processing process, not only suitable for large batches It also realizes high-precision monitoring of the shape and performance parameters of samples during processing.

Description

Technical field[0001]The present invention belongs to the field of laser precision machining and detection, and in particular to a depth learning-based femtosecond laser machining parameter confocal axial monitoring method.Background technique[0002]During the flying second laser processing, the processing material is different, the mechanism of the femtosecond pulse laser and the material is different, and the morphology and performance variations produced during the processing, the molecular structure, element ratio and charge of the sample under the action of pulsed laser The ions, etc., how to change the physical parameters and morphological parameters of the sample after processing, not only to ensure the key to machining accuracy, but also the important prerequisite for the study of femtosecond laser processing mechanism and improve processing technology.[0003]In the prior art, in the use of femtosecond laser processing, the sample is long and the efficiency is low, and it is n...

AI Technical Summary

Problems solved by technology

[0003] In the prior art, in the process of using femtosecond laser processing, the sample processing period is long and the efficiency is low, which cannot be applied in mass production applications

In addition, an additional auxiliary system is required to monitor the shape and performance parameters of the sample during processing with high precision, resulting in high production costs; at the same time, it limits the stability of femtosecond laser processing effects and cross-scale processing capabilities, and also restricts femtosecond laser processing. Research on processing mechanism and improvement of processing technology level

Images