Laser cladding, as an advanced green surface modification technology, has been widely applied in aerospace and other high-end fields due to its prominent advantages such as rapid alloy coating fabrication, cost-effectiveness, and high forming efficiency. The morphology of the cladding layer, a crucial characteristic indicator, encompasses key geometric parameters including width, height, depth of fusion, and contact angle, which directly determines the final forming quality and service performance of the components. However, traditional process optimization methods like single-variable analysis and orthogonal experiments suffer from significant efficiency bottlenecks. The trial-and-error parameter optimization process not only consumes substantial material and energy resources but also prolongs the research and development cycle. To address this challenge, establishing an accurate mathematical model for the cross-sectional morphology of single-track cladding has become the theoretical foundation for precise process control. Machine learning-based parameter optimization strategies which construct nonlinear mapping relationships between process parameters and forming quality, offer a promising way to break through the limitations of traditional methods, remarkably improving R&D efficiency and reducing experimental costs. Nevertheless, existing machine learning methods struggle to ensure optimal solutions for multiple objectives simultaneously in multi-dimensional optimization scenarios. In addition, the existing prediction models mostly focus on a single process parameter and fail to achieve comprehensive optimization of multi-dimensional parameters. Grey Wolf Optimizer (GWO), a robust multi-objective optimization algorithm inspired by the hunting behavior of grey wolves, excels in balancing global exploration and local exploitation, making it effective in overcoming this dilemma.
In this study, a polynomial regression model was established based on multiple process parameters to predict the width and height of the cladding layer. Meanwhile, a hybrid algorithm model combining GWO and Back Propagation Neural Network (GWO-BPNN) was proposed to forecast critical quality indicators including forming coefficient, contact angle, and dilution rate. Full-factor laser cladding experiments were conducted by depositing 316L alloy powder on 316L stainless steel substrates, followed by experimental validation and model inverse verification. The results demonstrated that the GWO-BPNN prediction model exhibited excellent performance in predicting key quality indicators. The average coefficient of determination (R2) reached 95.02%, a significant improvement of 12.4% compared with that of the traditional BPNN algorithm (R2 = 82.93%). Additionally, the relative error of 3D morphology reconstruction predicted by the regression model was controlled within 4.2%, verifying the practicality of the algorithm system in dynamic process parameter prediction and optimization. Both experimental and inverse verification confirmed the accuracy of the GWO-BPNN algorithm in predicting and optimizing various geometric characteristics of laser cladding layers. In conclusion, the overall prediction trend of the GWO-BPNN algorithm meets engineering tolerance requirements, providing a quantitative basis for the multi-dimensional optimization of cladding layer quality and validating the practicality and scientificity of the proposed evaluation indicators in industrial scenarios. This study is expected to effectively reduce test costs, improve production process stability and finished product quality, and provide new solutions for the intelligent development of laser cladding technology and also provides a theoretical basis and engineering practice basis for the intelligent upgrading of laser cladding process exploration.
Key words
laser cladding /
GWO-BPNN algorithm /
geometric characteristics /
regression analysis /
coefficient of determination
{{custom_sec.title}}
{{custom_sec.title}}
{{custom_sec.content}}
References
[1] LIU D Y, YANG X F, ZHAO A T, et al.Preparation of Nickel-Based Composite Coatings by Laser Cladding Technology: A Review[J]. The International Journal of Advanced Manufacturing Technology, 2024, 134(7): 3107-3137.
[2] DENG R, MAO M H, ZHAO C J, et al.A Review of Recent Advances in Integrated Laser Remelting and Laser Cladding Processes[J]. Metallurgical Research & Technology, 2024, 121(4): 402.
[3] BAI Q F, CHEN C, LI Q H, et al.Status of Research on Assisted Laser Cladding and Laser Cladding Posttreatment: A Review[J]. Physics of Metals and Metallography, 2024, 125(13): 1648-1663.
[4] LI Z Q, DU Y B, HU Y F.A Method for Predicting the Morphology of Single-Track Laser Cladding Layer Based on SO-LSSVR[J]. Materials Today Communications, 2024, 39: 108666.
[5] 刘丽兰, 李思聪, 豆卫涛, 等. 316L不锈钢表面激光熔覆Ni60合金涂层的工艺优化与性能研究[J]. 中国激光, 2024, 51(16): 1602207.
LIU L L, LI S C, DOU W T, et al.Process Optimization and Performance Analysis for Laser-Cladding Ni60 Alloy Coating on Surface of 316L Stainless Steel[J]. Chinese Journal of Lasers, 2024, 51(16): 1602207.
[6] 党霞. 激光表面改性技术在提升刀具耐磨性中的应用研究[J]. 农机使用与维修, 2025(2): 84-86.
DANG X.Research on the Application of Laser Surface Modification Technology in Enhancing Tool Wear Resistance[J]. Farm Machinery Using & Maintenance, 2025(2): 84-86.
[7] 孔令辉, 刘东, 张雷, 等. 水力机械抗空蚀涂层研究进展[J]. 人民黄河, 2023, 45(6): 157-162.
KONG L H, LIU D, ZHANG L, et al.Research Progress of Anti-Cavitation Coatings of Hydraulic Machinery[J]. Yellow River, 2023, 45(6): 157-162.
[8] PANT P, CHATTERJEE D.Prediction of Clad Characteristics Using ANN and Combined PSO-ANN Algorithms in Laser Metal Deposition Process[J]. Surfaces and Interfaces, 2020, 21: 100699.
[9] CHEN T, WU W N, LI W P, et al.Laser Cladding of Nanoparticle TiC Ceramic Powder: Effects of Process Parameters on the Quality Characteristics of the Coatings and Its Prediction Model[J]. Optics & Laser Technology, 2019, 116: 345-355.
[10] ILANLOU M, SHOJA RAZAVI R, NOUROLLAHI A, et al.Prediction of the Geometric Characteristics of the Laser Cladding of Inconel 718 on the Inconel 738 Substrate via Genetic Algorithm and Linear Regression[J]. Optics & Laser Technology, 2022, 156: 108507.
[11] SAINI N, SAHA S, JANGRA A, et al.Extractive Single Document Summarization Using Multi-Objective Optimization: Exploring Self-Organized Differential Evolution, Grey Wolf Optimizer and Water Cycle Algorithm[J]. Knowledge-Based Systems, 2019, 164: 45-67.
[12] 罗亮斌, 梁国星, 刘东刚, 等. 42CrMo钢表面激光熔覆钴基金刚石耐磨层组织及性能[J]. 表面技术, 2024, 53(5): 96-107.
LUO L B, LIANG G X, LIU D G, et al.Microstructure and Properties of Laser Cladding Co-Based Diamond Wear Resistant Layer on 42CrMo Steel Surface[J]. Surface Technology, 2024, 53(5): 96-107.
[13] 朱艳青. 车轴钢表面激光熔覆铁基合金涂层研究[J]. 电镀与精饰, 2019, 41(2): 17-22.
ZHU Y Q.Laser Cladding of Fe-Based Alloy Coating on Axle Steel Surface[J]. Plating & Finishing, 2019, 41(2): 17-22.
[14] 陈儒森, 吉小超, 张梦清, 等. 机器学习在激光熔覆涂层缺陷检测中的研究现状与进展[J]. 中国表面工程, 2024, 37(5): 112-137.
CHEN R S, JI X C, ZHANG M Q, et al.Progress and Research Status in Machine Learning for Defect Detection in Laser Cladding Coatings[J]. China Surface Engineering, 2024, 37(5): 112-137.
[15] 温海骏, 孟小玲, 许向川, 等. 基于神经网络和遗传算法的激光熔覆工艺参数多目标优化[J]. 应用激光, 2019, 39(5): 734-740.
WEN H J, MENG X L, XU X C, et al.Multi-Objective Optimization of Laser Cladding Process Parameters Based on Neural Network and Genetic Algorithm[J]. Applied Laser, 2019, 39(5): 734-740.
[16] RODRÍGUEZ-MOLINA A, VILLARREAL-CERVANTES M G, PANTOJA-GARCÍA J S, et al. Metaheuristic Adaptive Control Based on Polynomial Regression and Differential Evolution for Robotic Manipulators[J]. Applied Soft Computing, 2024, 151: 111116.
[17] LYU B H, LI J Z.Multiple-Model Polynomial Regression and Efficient Algorithms for Data Analysis[J]. Theoretical Computer Science, 2024, 1021: 114878.
[18] PUNEETH T, PANDA B K, NATH A K, et al.Effect of Dilution on the Properties of Laser Deposited AlCoCrFeNi High Entropy Alloy over SS316 Substrate[J]. Surface and Coatings Technology, 2025, 498: 131828.
[19] 范福杰. 激光熔覆718高温合金涂层的研究[D]. 兰州: 兰州理工大学, 2018.
FAN F J.Research on Laser Cladding Coating of 718 Superalloy[D]. Lanzhou: Lanzhou University of Technology, 2018.
[20] NAESSTROEM H, BRUECKNER F, KAPLAN A F H. Blown Powder Directed Energy Deposition on Various Substrate Conditions[J]. Journal of Manufacturing Processes, 2022, 73: 660-667.
[21] PATURI U M R, CHERUKU S, GEEREDDY S R. Process Modeling and Parameter Optimization of Surface Coatings Using Artificial Neural Networks (ANNs): State-of-the-Art Review[J]. Materials Today: Proceedings, 2021, 38: 2764-2774.
[22] JIANG F, FEI L Y, JIANG H, et al.Constitutive Model Research on the Hot Deformation Behavior of Ti6Al4V Alloy under Wide Temperatures[J]. Journal of Materials Research and Technology, 2023, 23: 1062-1074.
[23] 毛恺奕, 杜彦斌, 何国华, 等. 基于GWO-RFR的激光熔覆多道成形层形貌的预测方法[J]. 材料热处理学报, 2024, 45(2): 174-183.
MAO K Y, DU Y B, HE G H, et al.Prediction Method for Multi-Track Laser Cladding Layer Morphology Based on GWO-RFR[J]. Transactions of Materials and Heat Treatment, 2024, 45(2): 174-183.
[24] 周志杰. 20Cr13不锈钢表面激光熔覆15-5PH合金涂层工艺优化及性能研究[D]. 重庆: 重庆工商大学, 2022.
ZHOU Z J.Optimization and Performance of Laser Cladding 15-5PH Alloy Coating on 20Cr13 Stainless Steel Surface[D]. Chongqing: Chongqing Technology and Business University, 2022.
Funding
The National Key R&D Program of China (2024YFB3816500); Zhejiang Province "Pioneer" Research and Development Tackling Key Project (2023C01064); Zhejiang Province High-Level Talent Special Support Plan（2023R5210）; Longyou County Science and Technology Program Project（JHXM2023072）
PDF(15550 KB)
