基于机器学习的选区激光熔化形貌监测研究

金属选区激光熔化是一种通过激光逐层扫描金属粉末，将其熔结为零件的增材制造方法。因可提供几何形状复杂的定制化成型件，选区激光熔化在工业领域均有广泛的应用前景，但其成形质量的稳定性与一致性难以保证依然是阻碍该制造技术进一步发展难题。选区激光熔化形貌是一项重要的加工特征，一方面形貌研究有助于更好的理解加工过程中的物理现象，另一方面形貌监测是实现一系列成形质量控制与加工成本优化方法的基础。然而现有形貌监测面临图像特征选择、提取困难，数据处理速度慢等问题，因此亟需高效率、泛化性强的形貌标定算法与对应的监测方法。本文针对上述问题，通过研究选区激光熔化形貌随加工工艺的演变规律，提出了基于机器学习的熔池、熔道和熔面形貌监测方法，并对方法相关的形貌图像数据集、机器学习模型和对应算法展开研究。

本文总结了现有选区激光熔化形貌研究和监测中存在的问题和机器学习方法适用的难点，从成形面内熔池、熔道再到熔面，逐级递进地对选区激光熔化形貌进行图像采集和特征分析。结果发现熔池的特征尺寸在小步长增加输入脉冲激光能量时会产生均值无显著差异现象。熔道形貌受到激光功率等五种因素的影响，可分为过熔、稳定、欠熔三个状态。熔面的形貌会随输入激光功率增加，逐渐呈现欠熔、稳定与过熔三种特征，通过五个维度的粗糙度参数研究，发现熔面质量也随功率变化，但无单调线性关系。

为了实现加工过程中熔池状态变化的辨识，提出了基于卷积神经元网络的熔池形貌图像分类方法，建立了熔池形貌图像数据集，对比研究了八种卷积神经元网络模型的分类表现，使用神经元网络可视化方法对分类机理进行了研究。研究结果表明：卷积神经元网络能以最高96.6%的准确率对五个类别的熔池形貌图像分类，证明了方法的有效性。通用图像数据集上的预训练可以缩短模型的训练周期，提高分类精确度。可视化结果表明熔池区域对分类结果的贡献权重最大，证明了卷积神经元网络分类的可靠性，为该方法在熔池状态监测中的应用提供了有效的技术保障。

为了实现加工工艺参数的高效筛选，提出了基于深度学习目标检测的熔道形貌识别和分类方法，对三种熔道形貌标定建立了训练蒙版，建立了基于Single Shot MultiBox Detector(SSD)架构、以ResNet为主干网络的目标检测模型，研究了相关算法对模型表现的影响，对模型筛选工艺参数效果进行了分析。研究结果表明：将熔道标定为稳定、非稳定和过渡状态的标定方法有利于模型检测表现。目标检测模型的预训练是识别和分类熔道形貌的必要条件，同时特征融合和注意力算法显著提高了识别效果。结合高通量实验，通过熔道监测结果成功确定了316L不锈钢和纯铜粉加工条件下的适宜激光功率和扫描速度，确认了模型的泛化能力，验证了通过熔道形貌识别和分类筛选工艺参数的有效性。

为了实现加工过程中实时质量监测，提出了基于熔面形貌图像回归的成形密度预测方法，建立了图像回归模型，并针对模型训练对熔面形貌图像需求量较大的问题，提出了模拟熔面形貌图像的生成方法，最后对图像回归结果和密度预测效果进行了分析。结果表明：回归均值与标定真值最小误差为1%，根据熔面形貌图像回归均值的统计结果预测了成形部分块体材料的密度，预测结果与测量密度吻合度较高，验证了密度预测的可行性。在训练数据集中添加模拟图像后，相比于图像缺失时回归误差均值从65%降至16%，可使训练模型的密度预测误差降低0.05g/mm³，使模型逼近使用完备训练集时的密度预测效果。

[1] Beaman J, Bourell D L, Seepersad C, et al. Additive manufacturing review: early past to current practice[J]. Journal of Manufacturing Science and Engineering, 2020, 142(11): 110812-110832.
[2] Wohlers T, Campbell I, Diegel O, et al. Wohlers Report 2021: 3D Printing and additive manufacturing global state of the industry[M]. Fort Collins: Wohlers Associates, 2021: 1-334.
[3] Molnar M. The national network for manufacturing innovation[C]. Engineering Deans Public Policy Forum, 2014.
[4] Sargent J, John F. The Obama administration’s proposal to establish a national network for manufacturing innovation[J], Congressional Research Service, 2012: 1-21.
[5] Bonvillian W B. The rise of advanced manufacturing institutes in the United States[J]. The Next Production Revolution: Implications for Governments and Business, 2017: 55.
[6] Dilberoglu U M, Gharehpapagh B, Yaman U, et al. The role of additive manufacturing in the era of industry 4.0[J]. Procedia Manufacturing, 2017, 11: 545-554.
[7] Lemu H G. On opportunities and limitations of additive manufacturing technology for Industry 4.0 era[C]. International Workshop of Advanced Manufacturing and Automation. 2018: 106-113.
[8] 国家制造强国建设战略咨询委员会. 《中国制造2025》重点领域技术创新绿皮书[M]. 北京: 中国制造2025重点领域技术创新绿皮书, 2016.
[9] 伏欣. 国内增材制造(3D打印)技术发展现状与研究趋势[J]. 中国高新技术企业, 2016, (24): 2.
[10] 王红梅. 3D打印:先进制造领域的必争之地[J]. 杭州科技, 2013, 000(005): 35-38.
[11] Vafadar A, Guzzomi F, Rassau A, et al. Advances in metal additive manufacturing: a review of common processes, industrial applications, and current challenges[J]. Applied Sciences, 2021, 11(3): 1213.
[12] 王华明. 高性能金属构件增材制造技术开启国防制造新篇章[J]. 国防制造技术, 2013, (3): 3.
[13] 卢秉恒, 李涤尘. 增材制造(3D打印)技术发展[J]. 机械制造与自动化, 2013, 42(4): 4.
[14] 郭文文, 曹慧, 刘静. 3D打印技术在生物医学领域的应用[J]. 中国临床研究, 2016, 29(8): 3.
[15] Hebert R J. Metallurgical aspects of powder bed metal additive manufacturing[J]. Journal of Materials Science, 2016, 51(3): 1165-1175.
[16] Yan J, Zhou Y, Gu R, et al. A comprehensive study of steel powders (316L, H13, P20 and 18Ni300) for their selective laser melting additive manufacturing[J]. Metals, 2019, 9(1): 86.
[17] 史玉升, 鲁中良, 章文献, 等. 选择性激光熔化快速成形技术与装备[J]. 中国表面工程, 2006(z1): 150-153.
[18] 杨永强, 陈杰, 宋长辉, 等. 金属零件激光选区熔化技术的现状及进展[J]. 激光与光电子学进展, 2018, 55(1): 13.
[19] 全书海. 基于表面灰度图像的加工表面形貌分形特征研究[D]. 武汉: 武汉理工大学, 2003: 3-4.
[20] 林鑫, 黄卫东. 世界十大突破技术之首将在何处突破——谈金属增材制造研究[J]. 前沿科学, 2019, 13(4): 5.
[21] 卢秉恒. 智能制造与3D打印推动"中国制造2025"[J]. 高科技与产业化, 2018, (11): 4.
[22] 吴怀宇. 3D 打印: 三维智能数字化创造[M]. 北京: 电子工业出版社, 2017: 3-4.
[23] Das S. Physical aspects of process control in selective laser sintering of metals[J]. Advanced Engineering Materials, 2003, 5(10): 701-711.
[24] Abdulrahman K O, Akinlabi E T, Mahamood R M. Laser metal deposition technique: sustainability and environmental impact[J]. Procedia Manufacturing, 2018, 21: 109-116.
[25] Körner C. Additive manufacturing of metallic components by selective electron beam melting—a review[J]. International Materials Reviews, 2016, 61(5): 361-377.
[26] Debroy T, Wei H L, Zuback J, et al. Additive manufacturing of metallic components – process, structure and properties[J]. Progress in Materials Science, 2018, 92: pp. 112–224.
[27] Li J Z, Alkahari M R, Rosli N, et al. Review of wire arc additive manufacturing for 3D metal printing[J]. International Journal of Automation Technology, 2019, 13(3): 346-353.
[28] Varotsis A B. Introduction to metal 3D printing [EB/OL]. (2022)
[2022-2-1]. https://www.hubs.com/knowledge-base/introduction-metal-3d-printing/
[29] Papadakis L, Loizou A, Risse J, et al. A computational reduction model for appraising structural effects in selective laser melting manufacturing: a methodical model reduction proposed for time-efficient finite element analysis of larger components in selective laser melting[J]. Virtual and Physical Prototyping, 2014, 9(1): 17-25.
[30] Zhang L, Klemm D, Eckert J, et al. Manufacture by selective laser melting and mechanical behavior of a biomedical Ti–24Nb–4Zr–8Sn alloy[J]. Scripta Materialia, 2011, 65(1): 21-24.
[31] Vandenbroucke B, Kruth J P. Selective laser melting of biocompatible metals for rapid manufacturing of medical parts[J]. Rapid Prototyping Journal, 2007, 13(4): 196-203.
[32] Morgan R, Papworth A, Sutcliffe C, et al. High density net shape components by direct laser re-melting of single-phase powders[J]. Journal of Materials Science, 2002, 37(15): 3093-3100.
[33] Gunasekaran J, Sevvel P, Solomon I J. Metallic materials fabrication by selective laser melting: a review[J]. Materials Today: Proceedings, 2021, 37: 252-256.
[34] Heussen D, Meiners W. Green light for new 3D printing process[EB/OL]. (2017)
[2021-8-30]. https://www.ilt.fraunhofer.de/en/press/press-release-2017-08-30.html.
[35] Hönl R. World premiere at formnext: green laser from TRUMPF prints copper and gold [EB/OL]. (2018)
[2021-11-13]. https://www.trumpf.com/filestorage/ Formnext-Green-Laser-from-TRUMPF.pdf
[36] Gustmann T, Neves A, Kühn U, et al. Influence of processing parameters on the fabrication of a Cu-Al-Ni-Mn shape-memory alloy by selective laser melting[J]. Additive Manufacturing, 2016, 11: 23-31.
[37] Li R, Liu J, Shi Y, et al. Balling behavior of stainless steel and nickel powder during selective laser melting process[J]. The International Journal of Advanced Manufacturing Technology, 2012, 59(9): 1025-1035.
[38] Wang D, Wu S, Fu F, et al. Mechanisms and characteristics of spatter generation in SLM processing and its effect on the properties[J]. Materials & Design, 2017, 117: 121-130.
[39] Yuan P, Gu D. Molten pool behaviour and its physical mechanism during selective laser melting of TiC/AlSi10Mg nanocomposites: simulation and experiments[J]. Journal of Physics D: Applied Physics, 2015, 48(3): 035303.
[40] Purtonen T, Kalliosaari A, Salminen A. Monitoring and adaptive control of laser processes[J]. Physics Procedia, 2014, 56: 1218-1231.
[41] Yadav P, Rigo O, Arvieu C, et al. In situ monitoring systems of the SLM process: on the need to develop machine learning models for data processing[J]. Crystals, 2020, 10(6): 524.
[42] Mani M, Feng S, Brandon L, et al. Measurement science needs for real-time control of additive manufacturing powder-bed fusion processes[M]. Boca Raton: CRC Press, 2017: 22.
[43] Tanimura Y, Teague E, Scire F, et al. Graphical signatures for manufactured surfaces[J]. Journal of Tribology, 1982, 104(4): 533-537.
[44] Gusarov A, Kovalev E. Model of thermal conductivity in powder beds[J]. Physical Review B, 2009, 80(2): 024202.
[45] Thomas M, Baxter G J, Todd I. Normalised model-based processing diagrams for additive layer manufacture of engineering alloys[J]. Acta materialia, 2016, 108: 26-35.
[46] Körner C, Attar E, Heinl P. Mesoscopic simulation of selective beam melting processes[J]. Journal of Materials Processing Technology, 2011, 211(6): 978-987.
[47] Khairallah S A, Anderson A T, Rubenchik A, et al. Laser powder-bed fusion additive manufacturing: physics of complex melt flow and formation mechanisms of pores, spatter, and denudation zones[J]. Acta materialia, 2016, 108: 36-45.
[48] Gusarov A, Yadroitsev I, Bertrand P, et al. Heat transfer modelling and stability analysis of selective laser melting[J]. Applied surface science, 2007, 254(4): 975-979.
[49] Meier C, Penny R W, Zou Y, et al. Thermophysical phenomena in metal additive manufacturing by selective laser melting: Fundamentals, modeling, simulation and experimentation[J]. arXiv preprint, 2017: 170909510.
[50] Dai D, Gu D. Tailoring surface quality through mass and momentum transfer modeling using a volume of fluid method in selective laser melting of TiC/AlSi10Mg powder[J]. International Journal of Machine Tools and Manufacture, 2015, 88: 95-107.
[51] Zhang T, Li H, Liu S, et al. Evolution of molten pool during selective laser melting of Ti–6Al–4V[J]. Journal of Physics D: Applied Physics, 2018, 52(5): 055302.
[52] Mao Z, Zhang D Z, Wei P, et al. Manufacturing feasibility and forming properties of Cu-4Sn in selective laser melting[J]. Materials, 2017, 10(4): 333.
[53] Yadroitsev I, Smurov I. Surface morphology in selective laser melting of metal powders[J]. Physics Procedia, 2011, 12: 264-270.
[54] Furumoto T, Egashira K, Munekage K, et al. Experimental investigation of melt pool behaviour during selective laser melting by high speed imaging[J]. Cirp Annals, 2018, 67(1): 253-256.
[55] 林滨, 黄新雁, 魏莹, 等. 加工表面形貌测量理论、方法及评价[J]. 制造业自动化, 2006, 28(8): 5.
[56] Institution B S. Geometric product specifications surface texture: profile method: rules and procedures for the assessment of surface texture[M]. London: British Standards Institution, 1998.
[57] Iso E. 4287–Geometrical product specifications surface texture: definitions and surface texture parameters[M]. Geneva :International Organization for Standardization, 1997.
[58] Zecchino M. How to choose the correct stylus for any application [M]. Edina: Veeco Instruments Inc. 2005.
[59] Kerckhofs G, Pyka G, Moesen M, et al. High‐resolution microfocus X-ray computed tomography for 3D surface roughness measurements of additive manufactured porous materials[J]. Advanced Engineering Materials, 2013, 15(3): 153-158.
[60] Krishna A V. Towards topography characterization of additive manufacturing surfaces[D]. Sweden: Chalmers Tekniska Hogskola, 2020.
[61] Townsend A, Senin N, Blunt L, et al. Surface texture metrology for metal additive manufacturing: a review[J]. Precision Engineering, 2016, 46: 34-47.
[62] Boilot J P, Gelo P, Begin G. Adaptive welding by fiber optic thermographic sensing--an analysis of thermal and instrumental considerations[J]. Welding Journal, 1985, 64(7): 209-217.
[63] Krauss H, Eschey C, Zaeh M. Thermography for monitoring the selective laser melting process[C]. University of Texas at Austin: International Solid Freeform Fabrication Symposium, 2012: 2-3.
[64] Mazzucato F, Aversa A, Doglione R, et al. Influence of process parameters and deposition strategy on laser metal deposition of 316L powder[J]. Metals, 2019, 9(11): 1160.
[65] Kleszczynski S, Zur Jacobsmühlen J, Sehrt J, et al. Error detection in laser beam melting systems by high resolution imaging[C]. University of Texas at Austin: International Solid Freeform Fabrication Symposium, 2012:5.
[66] Everton S K, Hirsch M, Stravroulakis P, et al. Review of in-situ process monitoring and in-situ metrology for metal additive manufacturing[J]. Materials & Design, 2016, 95(Apr.): 431-445.
[67] Rombouts M, Kruth J-P, Froyen L, et al. Fundamentals of selective laser melting of alloyed steel powders[J]. Cirp Annals, 2006, 55(1): 187-192.
[68] Chivel Y, Smurov I. On-line temperature monitoring in selective laser sintering/melting[J]. Physics Procedia, 2010, 5: 515-521.
[69] 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): 1089-1101.
[70] Furumoto T, Alkahari M R, Ueda T, et al. Monitoring of laser consolidation process of metal powder with high speed video camera[J]. Physics Procedia, 2012, 39: 760-766.
[71] Yadroitsev I, Gusarov A, Yadroitsava I, et al. Single track formation in selective laser melting of metal powders[J]. Journal of Materials Processing Technology, 2010, 210(12): 1624-1631.
[72] Caltanissetta F, Grasso M, Petro S, et al. Characterization of in-situ measurements based on layerwise imaging in laser powder bed fusion[J]. Additive Manufacturing, 2018, 24: 183-199.
[73] Zur Jacobsmühlen J, Kleszczynski S, Witt G, et al. Elevated region area measurement for quantitative analysis of laser beam melting process stability[C]. International Solid Freeform Fabrication Symposium, 2015.
[74] 杨红平, 王斌. 机械加工表面形貌的分形特性表征[J]. 天水师范学院学报, 2011, 31(2): 3.
[75] 史立新. 车削加工表面的分形特性研究[J]. 农机化研究, 2002, (3): 3.
[76] Newell A, Shaw J C, Simon H A. Report on a general problem solving program[C]. Pittsburgh: IFIP congress, 256: 64.
[77] Dincbas M. A knowledge-based expert system for automatic analysis and synthesis in CAD[C]. IFIP congress, 1980.
[78] Yao B, Yang H. Constrained markov decision process modeling for sequential optimization of additive manufacturing build quality[J]. IEEE Access, 2018, 6: 54786-54794.
[79] Yao B, Imani F, Yang H. Markov decision process for image-guided additive manufacturing[J]. IEEE Robotics and Automation Letters, 2018, 3(4): 2792-2798.
[80] Shevchik S A, Kenel C, Leinenbach C, et al. Acoustic emission for in situ quality monitoring in additive manufacturing using spectral convolutional neural networks[J]. Additive Manufacturing, 2018, 21: 598-604.
[81] Kwon O, Kim H G, Ham M J, et al. A deep neural network for classification of melt-pool images in metal additive manufacturing[J]. Journal of Intelligent Manufacturing, 2020, 31(2): 375-386.
[82] Wu B, Ji X Y, Zhou J X, et al. In situ monitoring methods for selective laser melting additive manufacturing process based on images-a review[J]. China Foundry, 2021, 18(4): 265-285.
[83] Baumgartl H, Tomas J, Buettner R, et al. A deep learning-based model for defect detection in laser-powder bed fusion using in-situ thermographic monitoring[J]. Progress in Additive Manufacturing, 2020, 5(3): 277-285.
[84] Scime L, Beuth J. Anomaly detection and classification in a laser powder bed additive manufacturing process using a trained computer vision algorithm[J]. Additive Manufacturing, 2018, 19: 114-126.
[85] Scime L, Beuth J. Melt pool geometry and morphology variability for the Inconel 718 alloy in a laser powder bed fusion additive manufacturing process[J]. Additive Manufacturing, 2019, 29: 100830.
[86] Rajpurkar P, Zhang J, Lopyrev K, et al. Squad: 100,000+ questions for machine comprehension of text[J]. arXiv preprint, 2016:160605250.
[87] Wang A, Singh A, Michael J, et al. GLUE: A multi-task benchmark and analysis platform for natural language understanding[J]. arXiv preprint, 2018:180407461.
[88] Krizhevsky A, Hinton G. Learning multiple layers of features from tiny images[M/OL],2009
[2021-12-1]. http://doi=10.1.1.222.9220
[89] Krizhevsky A, Sutskever I, Hinton G E. Imagenet classification with deep convolutional neural networks[J]. Communications of the ACM, 2017, 60(6): 84-90.
[90] Lin T-Y, Maire M, Belongie S, et al. Microsoft coco: common objects in context[C]. European conference on computer vision, 2014: 740-755.
[91] Guyon I, Gunn S, Nikravesh M, et al. Feature extraction: foundations and applications[M]. Springer, 2008.
[92] Loussaief S, Abdelkrim A. Machine learning framework for image classification[C]. Technologies of Information and Telecommunications (SETIT), 2016: 58-61.
[93] Zou X. A review of object detection techniques[C]. International Conference on Smart Grid and Electrical Automation (ICSGEA), 2019: 251-254.
[94] Creswell A, White T, Dumoulin V, et al. Generative adversarial networks: an overview[J]. IEEE Signal Processing Magazine, 2018, 35(1): 53-65.
[95] Rumelhart D E, Hinton G E, Williams R J. Learning representations by back-propagating errors[J]. nature, 1986, 323(6088): 533-536.
[96] Glorot X, Bengio Y. Understanding the difficulty of training deep feedforward neural networks[C]. Proceedings of the thirteenth international conference on artificial intelligence and statistics, 2010: 249-256.
[97] Panwisawas C, Tang Y T, Reed R C. Metal 3D printing as a disruptive technology for superalloys[J]. Nature communications, 2020, 11(1): 1-4.
[98] Yang T, Liu T, Liao W, et al. The influence of process parameters on vertical surface roughness of the AlSi10Mg parts fabricated by selective laser melting[J]. Journal of Materials Processing Technology, 2019, 266: 26-36.
[99] Cao L, Yuan X. Study on the numerical simulation of the SLM molten pool dynamic behavior of a nickel-based superalloy on the workpiece scale[J]. Materials, 2019, 12(14): 2272.
[100] Liu J, Gu D-D, Chen H-Y, et al. Influence of substrate surface morphology on wetting behavior of tracks during selective laser melting of aluminum-based alloys[J]. Journal of Zhejiang University-SCIENCE A, 2018, 19(2): 111-121.
[101] Wang S, Liu Y, Shi W, et al. Research on high layer thickness fabricated of 316L by selective laser melting[J]. Materials, 2017, 10(9): 1055.
[102] Colopi M, Demir A G, Caprio L, et al. Limits and solutions in processing pure Cu via selective laser melting using a high-power single-mode fiber laser[J]. The International Journal of Advanced Manufacturing Technology, 2019, 104(5): 2473-2486.
[103] Ma C, Vadali M, Duffie N A, et al. Melt pool flow and surface evolution during pulsed laser micro polishing of Ti6Al4V[J]. Journal of Manufacturing Science and Engineering, 2013, 135(6).
[104] Dai W, Li J, Zhang W, et al. Evaluation of fluences and surface characteristics in laser polishing SKD 11 tool steel[J]. Journal of Materials Processing Technology, 2019, 273: 116241.
[105] Fox J C, Lopez F, Lane B M, et al. On the requirements for model-based thermal control of melt pool geometry in laser powder bed fusion additive manufacturing[C]. Salt Lake City: Proceedings of the Material Science & Technology Conference. 2016: 133-140.
[106] El N I, Murphy M J. What is machine learning?[M]. Machine learning in Radiation Oncology. Springer, 2015: 3-11.
[107] Luo W, Li Y, Urtasun R, et al. Understanding the effective receptive field in deep convolutional neural networks[J]. Advances in neural information processing systems, 2016, 29.
[108] He K, Zhang X, Ren S, et al. Deep residual learning for image recognition[C]. Proceedings of the IEEE conference on computer vision and pattern recognition, 2016: 770-778.
[109] Iandola F N, Han S, Moskewicz M W, et al. SqueezeNet: AlexNet-level accuracy with 50x fewer parameters and< 0.5 MB model size[J]. arXiv preprint, 2016:160207360.
[110] Vlachos M, Domeniconi C, Gunopulos D, et al. Non-linear dimensionality reduction techniques for classification and visualization[C]. Proceedings of the eighth ACM SIGKDD international conference on Knowledge discovery and data mining, 2002: 645-651.
[111] Bottou L. Stochastic gradient descent tricks[M]. Neural networks: Tricks of the trade. Springer, 2012: 421-436.
[112] Ruder S. An overview of gradient descent optimization algorithms[J]. arXiv preprint, 2016:160904747.
[113] Deng J, Dong W, Socher R, et al. Imagenet: A large-scale hierarchical image database[C]. IEEE conference on computer vision and pattern recognition, 2009: 248-255.
[114] Sun P, Feng W, Han R, et al. Optimizing network performance for distributed dnn training on gpu clusters: Imagenet/alexnet training in 1.5 minutes[J]. arXiv preprint, 2019:190206855.
[115] Jangapally T. Image Classification Using Network Inception-Architecture & Appications[J]. Journal For Innovative Development in Pharmaceutical and Technical Science, 2021, 4: 6-9.
[116] Yosinski J, Clune J, Nguyen A, et al. Understanding neural networks through deep visualization[J]. arXiv preprint, 2015:150606579.
[117] Zeiler M D, Fergus R. Visualizing and understanding convolutional networks[C]. European conference on computer vision. Springer, 2014: 818-833.
[118] Selvaraju R, Cogswell M, Das A, et al. Visual explanations from deep networks via gradient-based localization[C]. Proceedings of the IEEE International Conference on Computer Vision, 2017: 618-626.
[119] Shin M, Kim M, Kwon D-S. Baseline CNN structure analysis for facial expression recognition[C]. 25th IEEE international symposium on robot and human interactive communication, 2016: 724-729.
[120] Sarki R, Ahmed K, Wang H, et al. Image preprocessing in classification and identification of diabetic eye diseases[J]. Data Science and Engineering, 2021, 6(4): 455-471.
[121] Yang Z, Lu Y, Yeung H, et al. Investigation of deep learning for real-time melt pool classification in additive manufacturing[C]. Ieee 15th international conference on automation science and engineering, 2019: 640-647.
[122] Gao P, Wang Z, Zeng X. Effect of process parameters on morphology, sectional characteristics and crack sensitivity of Ti-40Al-9V-0.5 Y alloy single tracks produced by selective laser melting[J]. International Journal of Lightweight Materials and Manufacture, 2019, 2(4): 355-361.
[123] Sing S L, Wiria F E, Yeong W Y. Selective laser melting of titanium alloy with 50 wt% tantalum: effect of laser process parameters on part quality[J]. International Journal of Refractory Metals and Hard Materials, 2018, 77: 120-127.
[124] Dong Z, Liu Y, Wen W, et al. Effect of hatch spacing on melt pool and as-built quality during selective laser melting of stainless steel: modeling and experimental approaches[J]. Materials, 2019, 12(1): 50.
[125] Yadroitsev I, Smurov I. Selective laser melting technology: from the single laser melted track stability to 3D parts of complex shape[J]. Physics Procedia, 2010, 5: 551-560.
[126] Zhang B, Li Y, Bai Q. Defect formation mechanisms in selective laser melting: a review[J]. Chinese Journal of Mechanical Engineering, 2017, 30(3): 515-527.
[127] Hu Z, Nagarajan B, Song X, et al. Formation of SS316L single tracks in micro selective laser melting: surface, geometry, and defects[J]. Advances in Materials Science and Engineering, 2019, 2019.
[128] Installations T, Line L. Subjective video quality assessment methods for multimedia applications[J]. Networks, 1999, 910(37): 5.
[129] 周志华. 《机器学习》[M]. 中国民商, 2016, 03(No.21): 93-93.
[130] Liu W, Anguelov D, Erhan D, et al. Ssd: Single shot multibox detector[C]. European conference on computer vision, 2016: 21-37.
[131] Rothe R, Guillaumin M, Gool L V. Non-maximum suppression for object detection by passing messages between windows[C]. Asian conference on computer vision, 2014: 290-306.
[132] Li D, Yao A, Chen Q. PSConv: Squeezing Feature Pyramid into One Compact Poly-Scale Convolutional Layer[J]. arXiv preprint, 2020: 200706191.
[133] Cao J, Li Y, Sun M, et al. DO-Conv: Depthwise Over-parameterized Convolutional Layer[J]. arXiv preprint, 2020: 200612030.
[134] Vaswani A, Shazeer N, Parmar N, et al. Attention is all you need[C]. Advances in neural information processing systems, 2017: 5998-6008.
[135] Di W, Yongqiang Y, Xubin S, et al. Study on energy input and its influences on single-track, multi-track, and multi-layer in SLM[J]. The International Journal of Advanced Manufacturing Technology, 2012, 58(9): 1189-1199.
[136] Sadali M F, Hassan M Z, Ahmad F, et al. Influence of selective laser melting scanning speed parameter on the surface morphology, surface roughness, and micropores for manufactured Ti6Al4V parts[J]. Journal of materials research, 2020, 35(15): 2025-2035.
[137] Park H, Tran N, Nguyen D. Development of a predictive system for SLM product quality[C]. IOP Conference Series: Materials Science and Engineering, 2017(227): 012090.
[138] Radford A, Metz L, Chintala S. Unsupervised representation learning with deep convolutional generative adversarial networks[J]. arXiv preprint, 2015: 151106434.
[139] Baucher B, Chaudhary A B, Babu S S, et al. Defect characterization through automated laser track trace identification in SLM processes using laser profilometer data[J]. Journal of Materials Engineering and performance, 2019, 28(2): 717-727.