Email Alert  RSS
激光加工及增材制造技术专栏

基于机器学习的飞秒激光加工微坑阵列特征预测

展开
  • 北京工业大学材料与制造学部强场与超快光子学实验室

王 冰(1989-),助理研究员,工学博士,研究方向为激光智能制造,E-mail:wangbing@bjut.edu.cn;

宋海英(1979-),副研究员,工学博士,研究方向为激光精密加工, E-mail:hysong@bjut.edu.cn



收稿日期: 2023-02-01

  修回日期: 2023-03-20

  录用日期: 2023-04-15

  网络出版日期: 2023-10-15

Machine Learning-Assisted Feature Prediction of Micro Pit Arrays in Femtosecond Laser Processing

Expand
  • (Strong-Field and Ultrafast Photonics Lab, Faculty of Materials and Manufacturing, Beijing University of Technology, Beijing 100124, China)

Received date: 2023-02-01

  Revised date: 2023-03-20

  Accepted date: 2023-04-15

  Online published: 2023-10-15

摘要

微坑阵列结构具有耐磨损、抗腐蚀、提高生物相容性与抗菌性等特点,应用十分广泛。 飞秒激光独特的超快加工效应使其在高质量微坑加工中独具优势。 应用随机森林回归(RFR)算法和人工神经网络(ANN)算法对飞秒激光加工的微坑阵列几何形状和质量进行了预测,分析了激光加工参数对微坑的直径、深度和表面粗糙度(Ra)的影响。 通过均方根误差、确定系数以及平均绝对误差对RFR 与ANN 2 种模型的预测能力进行了评估。 结果显示:ANN 模型的整体预测准确率相比RFR 略高一些,R2值为0.81,直径、深度、粗糙度预测的R2分别为0.67、0.79、0.85。 利用数据增强方法对数据集进行了扩增,ANN 模型的准确率进一步提高,整体R2为0.91,直径、深度、粗糙度预测的R2分别为0.81、0.91、0.95。 研究结果表明,ANN 模型在飞秒激光加工微坑阵列的预测中相比RFR具有更优异的预测性能,且随着数据量的增加,这种优势更加明显,也进一步验证了ANN 模型的潜力。

本文引用格式

王泽林, 王冰, 宋海英, 刘世炳 . 基于机器学习的飞秒激光加工微坑阵列特征预测[J]. 材料保护, 2023 , 56(10) : 67 -73 . DOI: 10.16577/j.issn.1001-1560.2023.0236

Abstract

With the characteristics of wear resistance, corrosion resistance, improved biocompatibility and antibacterial properties, the micropits array structure has been widely used.femtosecond laser has unique advantages in high-quality micro pit processing, because of its unique ultra-fast processing effect.In this work, Random Forest Regression (RFR) algorithm and Artificial Neural Network (ANN) algorithm were applied to predict the geometry and quality of micro-pit arrays processed by femtosecond laser.Additionally, the effects of laser processing parameters on the diameter, depth and surface roughness (Ra) of micro-pits were analyzed.The predictive capabilities of the RFR and ANN models were evaluated through the root mean square error, coefficient of determination and mean absolute error.Results showed that the overall prediction accuracy of ANN model was slightly higher than RFR model.The R2 for ANN model was 0.81.For diameter, depth and surface roughness, the R2 was 0.67, 0.79 and 0.85, respectively.Data augmentation method was applied to augment the dataset, and the ANN model prediction accuracy was further improved after data augmentation.The overall R2 increased to 0.91.The R2 for diameter, depth and surface roughness was 0.81, 0.91 and 0.95, respectively.In general, the ANN model had better prediction performance than Random Forest in predicting micro pit arrays processed by femtosecond laser processing.As the amount of data increased, this advantage became more obvious, which further verified the potential of the ANN model.

参考文献

[1] 王耀南,陈铁健,贺振东,等.智能制造装备视觉检测控制方法综述[J].控制理论与应用,2015,32(3):273-286.WANG Y N,CHEN T J,HE Z D,et al.Review on the machine vision measurement and control technology for intelligent manufacturing equipment[J].Journal of Control Theory and Applications, 2015, 32(3): 273-286.

[2] CHOI H J, HUH D, JUN J, et al.A review on the fabrication and applications of sub-wavelength anti-reflective surfaces based on biomimetics [J].Applied Spectroscopy Reviews, 2019, 54(9): 719-735.

[3] BYUN J W, SHIN H S, KWON M H, et al.Surface texturing by micro ECM for friction reduction[J].International Journal of Precision Engineering and Manufacturing, 2010,11: 747-753.

[4] WANG G, WAN Y, LIU Z.Construction of Complex Structures Containing Micro-Pits and Nano-Pits on the Surface of Titanium for Cytocompatibility Improvement[J].Materials.2019, 12(17): 2 820.

[5] 王文中,黄志祥,沈 殿,等.圆柱形表面微坑阵列对点接触润滑摩擦性能的影响[J].摩擦学学报,2012,32(4):371-376.WANG W Z, HUANG Z X, SHEN D, et al.Effect of Patterned Cylindrical Dimple Array on the Tribological Performance of Lubricated Point-Contacts[J].Tribology, 2012, 32(4):371-376.

[6] LIN Y, HAN J, CAI M, et al.Durable and robust transparent superhydrophobic glass surfaces fabricated by a femtosecond laser with exceptional water repellency and thermostability[J].Journal of Materials Chemistry A, 2018, 6(19): 9 049-9 056.

[7] 顾 丰.电火花微小孔加工工艺参数优化及建模的研究[D].大连:大连理工大学, 2005.GU F.The study on parameter optimization and model of electric discharge machining (EDM) micro-and-small holes[D].Dalian: Dalian University of Technology, 2005.

[8] CAPANELLI S L, BONSERIO C.An artificial neural network approach for the control of the laser milling process[J].The International Journal of Advanced Manufacturing Technology, 2012, 66:1 777-1 784.

[9] 白基成,黄 河,郭永丰,等.高速电火花小孔加工电极损耗预测技术的研究[C]/ /2007 年中国机械工程学会年会论文汇编.长沙:中国机械工程学会特种加工分会, 2007.BAI J C,HUANG H,GUO Y F,et al.Prediction technique of high speed EDM-MILL electrode wear[C]/ /2007 Chinese Mechanical Engineering Society Conference Proceedings.Changsha: Special Processing Branch of Chinese Mechanical Engineering Society, 2007.

[10] DINAHARAN I, PALANIVEL R, MURUGAN N, LAUBSCHER R F.Application of artificial neural network in predicting the wear rate of copper surface composites produced using friction stir processing[J].Australian Journal of Mechanical Engineering, 2020, 13:1-12.

[11] ZHANG Z, LIU S, ZHANG Y, et al., Optimization of lowpower femtosecond laser trepan drilling by machine learning and a high-throughput multi-objective genetic algorithm[J].Optics and Laser Technology, 2022, 148:107688.

[12] SENTHIL KANNAN V, LENIN K, SUJITH KUMAR J.A parametric study on laser square hole machining of AA7475/SiC/ZrSiO4 composites[J].Materials Today: Proceedings,2022, 68:1 375-1 380.

[13] RAJESH N, GURU MAHESH G, VENKATARAMAIAH P,Prediction of taper angle in laser drilling of S32750 using integrated ANFIS[J].Materials Today: Proceedings, 2022,62:6 009-6 017.

[14] WANG B, WANG P, SONG J ,et al.A hybrid machine learning approach to determine the optimal processing window in femtosecond laser-induced periodic nanostructures[J].Journal of Materials Processing Technology, 2022(308):117-124.

[15] LEO B.Random forests[J].Machine Learning, 2001, 45(1): 5-32.

[16] TIBSHIRANI R.Bias, variance and prediction error for classification rules[M].Toronto: University of Toronto, 1996.

[17] STROBL C, BOULESTEIX A,KNEIB T, et al.Conditional variable importance for random forests[J].BMC Bioinformatics, 2008, 9(1): 307.

[18] LEK S,PARK Y S,Artificial Neural Networks,in Encyclopedia of Ecology[M].Oxford: Academic Press,2008.

[19] 刑松龄,刘 磊.飞秒激光参数对石英玻璃微孔加工的影响[J].中国激光,2015,42(4):264-269.XING S L, LIU L.Effects of femtosecond laser parameters on hole drilling of silica glass[J].Chinese Journal of Lasers, 2015, 42 (4): 264-269.

[20] SHARIFZADEH M, SIKINIOTI-LOCK A, SHAH N, Machine-learning methods for integrated renewable power generation:A comparative study of artificial neural networks,support vector regression, and Gaussian Process Regression[J].Renewable and Sustainable Energy Reviews,2019,108:513-538.

[21] DESAI S, OUARDA T B M J, Regional hydrological frequency analysis at ungauged sites with random forest regression[J].Journal of Hydrology, 2021, 594:125861.

文章导航

/