Process monitoring and quality identification of Powder Bed Fusion (PBF) are the key technologies to ensure its manufacturing quality. Among the monitoring signals, the molten pool radiation intensity signal contains rich molten pool characteristic information, but the direct connection between the monitoring signal and the material science phenomenon is not clear, and it is suitable to use machine learning algorithms to carry out in-depth research. Through 316L stainless steel forming experiments, different process parameters (partial deviation from the process window) are set, and the emission intensity signal of the molten pool is collected during the forming process. The signal data are preprocessed through data segmentation, feature extraction and feature selection, and a data set for machine learning is constructed. Twenty-one different machine learning algorithms are used. The molten pool emission intensity data are firstly classified according to the process parameters (such as high, medium, and low laser power). The trained algorithms can identify abnormalities (abnormal emission intensity, representative laser power and scanning speed deviate from the best state). Secondly, the emission intensity data of the molten pool are classified according to the quality (density, surface roughness) of the final formed block, and the quality can be identified in actual production through the trained algorithms. The results show that for the identification of abnormalities of the emission intensity of the molten pool, the classification effect of the secondary support vector machine algorithm is the best, with an accuracy of over 96.4%. For prediction of density and surface roughness, the prediction results show different distributions due to the complex relationship between the sample quality and the data sample, but the prediction accuracy has reached more than 96.0%.
[1] LI D C, LU Z L, TIAN X Y, et al. Additive manufacturing-Revolutionary technology for leading aerospace manufacturing[J]. Acta Aeronautica et Astronautica Sinica, 2022, 43(4): 525387 (in Chinese). 李涤尘, 鲁中良, 田小永, 等. 增材制造: 面向航空航天制造的变革性技术[J]. 航空学报, 2022, 43(4): 525387.
[2] HERZOG D, SEYDA V, WYCISK E, et al. Additive manufacturing of metals[J]. Acta Materialia, 2016, 117: 371-392.
[3] LU B H, LI D C, TIAN X Y. Development trends in additive manufacturing and 3D printing[J]. Engineering, 2015, 1(1): 85-89.
[4] GU D D, MA C L, XIA M J, et al. A multiscale understanding of the thermodynamic and kinetic mechanisms of laser additive manufacturing[J]. Engineering, 2017, 3(5): 675-684.
[5] KING W E, ANDERSON A T, FERENCZ R M, et al. Laser powder bed fusion additive manufacturing of metals: Physics, computational, and materials challenges[J]. Applied Physics Reviews, 2015, 2(4): 041304.
[6] TAPIA G, ELWANY A. A review on process monitoring and control in metal-based additive manufacturing[J]. Journal of Manufacturing Science and Engineering, 2014, 136(6): 060801.
[7] CRAEGHS T, BECHMANN F, BERUMEN S, et al. Feedback control of layerwise laser melting using optical sensors[J]. Physics Procedia, 2010, 5: 505-514.
[8] ZUR JACOBSMVHLEN J, KLESZCZYNSKI S, SCHNEIDER D, et al. High resolution imaging for inspection of laser beam melting systems[C]//2013 IEEE International Instrumentation and Measurement Technology Conference (I2MTC). Piscataway: IEEE Press, 2013: 707-712.
[9] GRASSO M, COLOSIMO B M. A statistical learning method for image-based monitoring of the plume signature in laser powder bed fusion[J]. Robotics and Computer-Integrated Manufacturing, 2019, 57: 103-115.
[10] GRASSO M, DEMIR A G, PREVITALI B, et al. In situ monitoring of selective laser melting of zinc powder via infrared imaging of the process plume[J]. Robotics and Computer-Integrated Manufacturing, 2018, 49: 229-239.
[11] ANUSUYA M A, KATTI S K. Speech recognition by machine: A review[DB/OL]. arXiv preprint: 1001.2267, 2010.
[12] CAGGIANO A, ZHANG J J, ALFIERI V, et al. Machine learning-based image processing for on-line defect recognition in additive manufacturing[J]. CIRP Annals, 2019, 68(1): 451-454.
[13] GOBERT C, REUTZEL E W, PETRICH J, et al. Application of supervised machine learning for defect detection during metallic powder bed fusion additive manufacturing using high resolution imaging[J]. Additive Manufacturing, 2018, 21: 517-528.
[14] SCIME L, BEUTH J. Using machine learning to identify in situ melt pool signatures indicative of flaw formation in a laser powder bed fusion additive manufacturing process[J]. Additive Manufacturing, 2019, 25: 151-165.
[15] WANG C, TAN X P, TOR S B, et al. Machine learning in additive manufacturing: State-of-the-art and perspectives[J]. Additive Manufacturing, 2020, 36: 101538.
[16] CAO L C, ZHOU Q, HAN Y F, et al. Review on intelligent monitoring of defects and process control of selective laser melting additive manufacturing[J]. Acta Aeronautica et Astronautica Sinica, 2021, 42(10): 524790 (in Chinese). 曹龙超, 周奇, 韩远飞, 等. 激光选区熔化增材制造缺陷智能监测与过程控制综述[J]. 航空学报, 2021, 42(10): 524790.
[17] CLIJSTERS S, CRAEGHS T, BULS S, et al. In situ quality control of the selective laser melting process using a high-speed, real-time melt pool monitoring system[J]. The International Journal of Advanced Manufacturing Technology, 2014, 75(5-8): 1089-1101.
[18] MONTAZERI M, YAVARI R, RAO P, et al. In-process monitoring of material cross-contamination defects in laser powder bed fusion[J]. Journal of Manufacturing Science and Engineering, 2018, 140(11): 111001.
[19] COECK S, BISHT M, PLAS J, et al. Prediction of lack of fusion porosity in selective laser melting based on melt pool monitoring data[J]. Additive Manufacturing, 2019, 25: 347-356.
[20] FERRI F, PUDIL P, HATEF M, et al. Comparative study of techniques for large-scale feature selection[J]. Machine Intelligence and Pattern, 1994, 16: 403-413.
[21] DASH M, LIU H. Feature selection for classification[J]. Intelligent Data Analysis, 1997, 1(1-4): 131-156.
[22] LI J D, CHENG K W, WANG S H, et al. Feature selection[J]. ACM Computing Surveys, 2018, 50(6): 1-45.
[23] PENG H C, LONG F H, DING C. Feature selection based on mutual information criteria of max-dependency, max-relevance, and min-redundancy[J]. IEEE Transactions on Pattern Analysis and Machine Intelligence, 2005, 27(8): 1226-1238.
[24] DARBELLAY G A, VAJDA I. Estimation of the information by an adaptive partitioning of the observation space[J]. IEEE Transactions on Information Theory, 1999, 45(4): 1315-1321.
[25] GUOLIN K, QI M, THOMAS F, et al. Lightgbm: A highly efficient gradient boosting decision tree[J]. Advances in Neural Information Processing Systems, 2017, 30: 3146-3154.
[26] CHANG C C, LIN C J. LIBSVM: A library for support vector machines[J]. ACM Transactions on Intelligent Systems and Technology, 2011, 2(3): 27.
[27] KOBY C, YORAM S. On the algorithmic implementation of multiclass kernel-based vector machines[J]. Journal of Machine Learning Research, 2001, 2: 265-292.
[28] BOTTOU L. Stochastic gradient descent tricks[M]//Lecture Notes in Computer Science. Berlin: Springer, 2012: 421-436.
[29] SEBASTIAN R. An overview of gradient descent optimization algorithms[DB/OL]. arXiv preprint: 1609.04747, 2016.
[30] NITESH V C, NATHAILE J, ALEKSANDAR K. Special issue on learning from imbalanced data sets[J]. ACM SIGKDD Explorations Newsletter, 2004, 6(1): 1-6.
[31] HAN H, JIANG X Q. Overcome support vector machine diagnosis overfitting[J]. Cancer Informatics, 2014, 13(S1): CIN. S13875.
[32] POZZOLO A D, CAELEN O, JOHNSON R A, et al. Calibrating probability with undersampling for unbalanced classification[C]//2015 IEEE Symposium Series on Computational Intelligence. Piscataway: IEEE Press, 2015: 159-166.
[33] TYAGI S, MITTAL S. Sampling approaches for imbalanced data classification problem in machine learning[M]//Lecture Notes in Electrical Engineering. Cham: Springer International Publishing, 2019: 209-221.
[34] EITRICH T, LANG B. Efficient optimization of support vector machine learning parameters for unbalanced datasets[J]. Journal of Computational and Applied Mathematics, 2006, 196(2): 425-436.
CJA