Modeling pulsed laser micromachining of micro geometries using machine-learning techniques
A wide range of opportunities are emerging in the micro-system technology sector for laser micro-machining systems, because they are capable of processing various types of materials with micro-scale precision. However, few process datasets and machine-learning techniques are optimized for this industrial task. This study describes the process parameters of micro-laser milling and their influence on the final features of the microshapes that are produced. It also identifies the most accurate machine-learning technique for the modelization of this multivariable process. It examines the capabilities of laser micro-machining by performing experiments on hardened steel with a pulsed Nd:YAG laser. Arrays of micro-channels were manufactured using various scanning speeds, pulse intensities and pulse frequencies. The results are presented in terms of the main industrial requirements for any manufactured good: dimensional accuracy (in our case, depth and width of the channels), surface roughness and material removal rate (which is a measure of the productivity of the process). Different machine-learning techniques were then tested on the datasets to build highly accurate models for each output variable. The selected techniques were: k-Nearest Neighbours, neural networks, decision trees and linear regression models. Our analysis of the correlation coefficients and the mean absolute error of all the generated models show that neural networks are better at modelling channel depth and that decision trees are better at modelling material removal rate; both techniques were similar for width and surface roughness. In general, these two techniques show better accuracy than the other two models. The work concludes that decision trees should be used, if information on input parameter relations is sought, while neural networks are suitable when the dimensional accuracy of the workpiece is the main industrial requirement. Extensive datasets are necessary for this industrial task, to provide reliable AI models due to the high rates of noise, especially for some outputs such as roughness.
KeywordsMachine learning-techniques Laser process Process parameters
This study was partially supported through grants from the European Commission project IREBID (FP7-PEOPLE-2009-IRSES-247476) and the Spanish Science and Innovation Minister project TECNIPLAD (DPI2009-09852).
- Desai, C. K., & Shaikh, A. (2012). Prediction of depth of cut for single-pass laser micro-milling process using semi-analytical, ANN and GP approaches. International Journal of Advanced Manufacturing Technology, 60(9–12), 865–882.Google Scholar
- Díez-Pastor, J. F., Bustillo, A., Quintana, G., & García-Osorio, C. (2012). Boosting projections to improve surface roughness prediction in high-torque milling operations. Soft Computing, 16(8), 1427–1437. Google Scholar
- Grzenda, M., Bustillo, A., Quintana, G., & Ciurana, J. (2012a). Improvement of surface roughness models for face milling operations through dimensionality reduction. Integrated Computer-Aided Engineering, 19(2), 179–197.Google Scholar