[1]张志华,侯晓磊*,杨 茂.doi: 10.3969/j.issn.1001-3849.2025.09.011基于改进神经网络的纳米镀层显微硬度预测方法[J].电镀与精饰,2025,(09):75-82.
　Zhang Zhihua,Hou Xiaolei *,Yang Mao.Prediction method of microhardness for nano coating based on improved neural network[J].Plating & Finishing,2025,(09):75-82.
点击复制

doi: 10.3969/j.issn.1001-3849.2025.09.011基于改进神经网络的纳米镀层显微硬度预测方法()

《电镀与精饰》[ISSN:1001-3849/CN:12-1096/TG]

卷:
期数:
2025年09
页码:
75-82
栏目:
出版日期:
2025-09-30

文章信息/Info

Title:
Prediction method of microhardness for nano coating based on improved neural network
作者:
张志华1; 侯晓磊1*; 杨 茂2
(1. 郑州工业应用技术学院 信息工程学院,河南 新郑 451150 ;2. 河南工业大学 经济贸易学院,河南 郑州 450000)
Author(s):
Zhang Zhihua1;  Hou Xiaolei 1*;  Yang Mao2
(1. Institute of Information Engineering, School of Zhengzhou Industrial Technology, Xinzheng 451150, China; 2. School of Economics and Trade, Henan University of Technology, Zhengzhou 450000, China)
关键词:
RBF神经网络; 纳米显微镀层; 因子分析法; 硬度预测; 鹈鹕优化算法
Keywords:
RBF neural network;  nano micro coating;  factor analysis method;  hardness prediction;  pelican optimization algorithm
分类号:
TG156.21
文献标志码:
A
摘要:
纳米镀层显微硬度预测过程中，纳米镀层显微硬度受材料成分、晶体结构、制备工艺参数等多种因素影响，难以精确量化各因素对硬度的单独贡献，导致建立准确显微硬度预测模型的难度增加。为此，提出基于改进神经网络的纳米镀层显微硬度预测方法。本文选用纯度为99.99 wt.%、直径为3 mm的镍棒作为阳极材料，选用Q235钢(尺寸为20 mm×10 mm×2 mm)作为阴极基体材料,并掺入不同浓度的SiC纳米粒子，通过磁场-喷射电沉积法来制备Ni-W-SiC纳米复合镀层试件。采用因子分析法选取纳米镀层显微硬度的关键影响因子，将其作为预测数据，有效量化各因素对硬度的贡献；通过模糊聚类算法对上述数据展开聚类处理，构建预测数据集；引入鹈鹕优化算法优化径向基函数(radial basis function, RBF)神经网络权值，实现神经网络的改进，建立精准的纳米镀层显微硬度预测模型，将预测数据集输入RBF神经网络模型中，输出纳米镀层显微硬度预测结果。结果表明，经过优化后RBF神经网络性能有所提升，具有较高的预测精度，且SiC纳米粒子浓度为8 g/L时，纳米镀层显微硬度可达最大，显微硬度为360 MPa。
Abstract:
In the process of predicting microhardness of nano coatings, the microhardness is influenced by various factors such as material composition, crystal structure, and preparation process parameters, making it difficult to accurately quantify the individual contributions of each factor to hardness, which increases the difficulty of establishing an accurate microhardness prediction model. Therefore, a microhardness prediction method for nano coatings based on improved neural networks is proposed. This article uses nickel rod with 99.99 wt.% purity and 3 mm diameter as the anode material, Q235 steel (size 20 mm×10 mm×2 mm) as the cathode matrix material, and different concentrations of SiC nanoparticles are doped. Ni-W-SiC nanocomposite coating specimens are prepared by magnetic field spray electrodeposition method. Using factor analysis to select key influencing factors on the microhardness of nano coatings, and using them as predictive data to effectively quantify the contribution of each factor to hardness; Cluster the above data using fuzzy clustering algorithm to construct a prediction dataset; Introducing the pelican optimization algorithm to optimize the weights of the RBF neural network, achieving improvements in the neural network, establishing an accurate prediction model for the microhardness of nano coatings, inputting the predicted dataset into the radial basis function (RBF) neural network model, and outputting the prediction results for the microhardness of nano coatings. The experimental results show that the performance of the optimized RBF neural network has been improved, with high prediction accuracy. When the concentration of SiC nanoparticles is 8 g/L, the microhardness of the nano coating can reach its maximum, with a microhardness of 360 MPa

参考文献/References:

[1].任鑫, 张雨辰, 田佳茹, 等. 不同方式电沉积Ni-纳米TiC复合镀层的微观结构与性能[J]. 复合材料学报, 2022, 39(8): 4093-4101.
[2].Lamichaney S, Mandal R, Mondal N, et al. Impact of coating variables on micro-hardness in Ni-W-P electroless coatings: experimental assessment, modelling and optimization[J]. Journal of The Institution of Engineers (India): Series C, 2024, 105(3): 595-606.
[3].陈辉, 傅作华, 陈得良, 等. 基于维氏压入理论模型的材料塑性参数与硬度一体化预测方法[J]. 机械工程学报, 2023, 59(8): 132-141.
[4].石明, 汪舟, 甘进, 等. 基于GA-BP神经网络的喷丸样品表层硬度预测模型[J]. 表面技术, 2022, 51(1): 332-338, 357.
[5].Kumar V, Verma R, Sharma V S, et al. Investigations into lattice strain and hardness of WC-10Co-4Cr coating on laser surface textured 420 stainless steel[J]. Lasers in Engineering (Old City Publishing), 2022, 53(4):231-252.
[6].梁月清, 刘会来, 崔康平, 等. 基于三维荧光光谱-平行因子分析法的工业园区污水溶解性有机物溯源与归趋[J]. 环境工程学报, 2022, 16(4): 1238-1247.
[7].阎莹, 王虹婷, 万利, 等. 基于因子分析与熵值法的不同隧道侧向宽度下驾驶负荷模型[J]. 中国公路学报, 2023, 36(2): 190-202.
[8].魏文军, 李政, 武晓春, 等. 基于模糊聚类算法的S700K型电动转辙机运行状态评估[J]. 铁道学报, 2022, 44(4): 74-81.
[9].乔彩彩, 吴成茂, 李昌兴, 等. 结合隶属度与像素交替引导滤波的鲁棒模糊聚类算法[J]. 计算机工程, 2022, 48(8): 224-233.
[10].化春键, 张爱榕, 蒋毅, 等. 基于改进模糊C均值聚类算法的草坪杂草识别[J]. 华南农业大学学报, 2022, 43(3): 107-115.
[11].戴世杰, 李哲明. 基于融合模糊聚类算法的异常心电多频段弱信号快速捕捉方法[J]. 现代电子技术, 2022, 45(21): 62-66.
[12].杨旭红, 陈阳, 贾巍, 等. 基于RBF神经网络的电压外环滑模控制的Vienna整流器[J]. 电力系统保护与控制, 2022, 50(18): 103-115.
[13].刘宜成, 熊宇航, 杨海鑫. 基于RBF神经网络的多关节机器人固定时间滑模控制[J]. 控制与决策, 2022, 37(11): 2790-2798.
[14].左锋琴, 张达敏, 何庆, 等. 融合无迹sigma点变异和交叉反向的鹈鹕优化算法[J]. 计算机科学与探索, 2024, 18(11): 2954-2968.
[15].马愈昭, 朱庆啸, 吕其明, 等. 基于POA-VMD和MPE的光纤周界入侵信号降噪方法[J]. 激光与光电子学进展, 2024, 61(17): 141-150.
[16].付传起, 林永威, 黄亚忠, 等. Ni-氧化石墨烯纳米复合镀层的力学及抗腐蚀性能研究[J]. 表面技术, 2023, 52(8): 290-300.
[17].王情情, 刘战强, 程延海, 等. 基于多尺度晶粒细化演变的TC4加工表层硬度预测[J]. 华南理工大学学报, 2023, 51(2): 35-46.
[18].郭明忠, 谭世光, 任爽, 等. Ni-P/纳米SiC复合镀层制备及性能研究[J]. 广州化工, 2024, 52(2): 97-98.
[19].邱有春, 赵立平. 基于改进神经网络的机身镀层抗冲击性能预测[J]. 兵器材料科学与工程, 2022, 45(5): 148-153.
[20].夏运生, 白鑫, 马三蕊. 改进BP神经网络的混凝土构件承载力预测仿真[J]. 计算机仿真, 2024, 41(4): 436-440.

相似文献/References:

[1]邢 翀*,王坤昊.基于BP和RBF神经网络预测硬质阳极氧化膜的硬度[J].电镀与精饰,2021,(7):25.[doi:10.3969/j.issn.1001-3849.2021.07.005]
　XING Chong*,WANG Kunhao.Prediction of Hardness of Hard Anodic Oxidation Film Based on BP and RBF Neural Network[J].Plating & Finishing,2021,(09):25.[doi:10.3969/j.issn.1001-3849.2021.07.005]

更新日期/Last Update: 2025-09-11