基于GA-BP神经网络的镍基合金熔覆涂层形貌预测

刘干成; 黄博

应用激光, 2018, 38 (4): 527, 网络出版: 2018-10-06

基于GA-BP神经网络的镍基合金熔覆涂层形貌预测

Prediction of Ni-based Alloy Cladding Coatings Topography Based on GA-BP Neural Network

论文大纲

刘干成 ^*黄博

作者单位

湖北工业大学机械工程学院, 湖北武汉 430068

激光熔覆正交实验方法遗传算法 BP神经网络 laser cladding orthogonal experiment method genetic algorithm BP neural network

摘要

针对镍基合金在熔覆过程中成形质量难以控制的问题, 通过建立3因素、4水平的正交试验参数表, 在完成22组实验样本数据的基础上, 建立经遗传算法优化的误差反向传播的神经网络模型, 通过激光加工工艺参数预测熔覆层宏观形貌, 形成两者之间的高度映射关系。采用极差分析法找出工艺参数对熔覆层宏观质量影响的大小关系并确定最佳工艺参数, 为了适用于大面积熔覆的需要, 引入熔覆层搭接率作为新的工艺参数来优化模型算法, 并通过宏观质量分析确定适宜的扫描间距, 利用5组不同搭接率的样本对模型的精度进行检验。实验结果表明, 经过优化后GA-BP模型具有很高的预测精度, 预测结果与测试样本之间的平均相对误差为3.951%, 验证了该模型在理论与实际中的可行性, 对提高镍基合金熔覆涂层的产品质量具有重要义。

Abstract

To solve the problem that the forming quality of Ni-based alloys is difficult to control during cladding. Via establishing a 3 factors and 4 levels orthogonal test parameter table and on the basis of completing 22 sets of experimental sample data, build an error reverse propagation neural network model which is optimized by genetic algorithm. Predicting the macroscopic morphology of the cladding layer by laser processing parameters, form a high mapping relationship between the two. Optimum parameters can be attained by finding out the influence of process parameters on the macroscopic quality of the cladding layer using range analysis method. To meet the needs of large area cladding, introduce the cladding layer overlap rate as a new process parameter to optimize the model algorithm, and the appropriate scanning interval is determined by the macro-quality analysis. Using 5 sets of samples with different overlap rates to test the accuracy of the model. The experimental results show that the optimized GA-BP model has a high prediction accuracy, and the average relative error between the predicted results and the test samples is 3.951%, which verifies the feasibility of the model in theory and practice, and is of great significance for improving the product quality of nickel-base alloy cladding coatings.

参考文献

[1] 杨宁, 杨帆.激光熔覆工艺参数对熔覆层质量的影响［J］.热处理技术与装备, 2013, 31(4): 17-19.

YANG N, YANG F.Influence of laser cladding process parameters on cladding layer quality［J］.Heat Treatment Technology and Equipment, 2010, 31(4): 17-19.

[2] 李华玲, 王茂才, 赵吉宾, 等.铝合金Nd:YAG脉冲激光SiC复合化涂层组织特征［J］.中国激光, 2004, 31(6): 765-768.

LI H, WANG M C, ZHAO J B, et al.Microstructure characteristics of aluminum alloy Nd:YAG pulsed laser SiC composite coating［J］.Chinese Journal of Lasers, 2004, 31(6): 765-768.

[3] 关泰红, 高勃, 吕晓卫, 等.激光快速成形工艺参数对生物陶瓷复合涂层物相组成的影响［J］.中国激光, 2009, 36(10): 2717-2721.

GUAN T H, GAO B, LV X W, et al.Effects of laser rapid forming process parameters on phase composition of bioceramic composite coatings［J］.Chinese Journal of Lasers, 2009, 36(10): 2717-2721.

[4] 张霜银, 林鑫, 陈静, 等.工艺参数对激光快速成形TC4钛合金组织及成形质量的影响［J］.稀有金属材料与工程, 2007, 36(10): 1839-1843.

ZHANG S Y, LIN X, CHEN J, et al.Effect of process parameters on microstructure and forming quality of laser rapid forming TC4 titanium alloy［J］.Rare Metal Materials and Engineering, 2007, 36(10): 1839-1843.

[5] 夏仁波, 刘伟军, 王越超.影响激光直接制造金属零件精度的因素及其闭环控制研究进展［J］.机械科学与技术, 2004, 23(9): 1085-1089.

XIA R B, LIU W J, WANG Y C.Research progress on factors affecting the precision of direct laser manufacturing of metal parts and its closed-loop control［J］.Mechanical Science and Technology for Aerospace Engineering, 2004, 23(9): 1085-1089.

[6] OCELK V, OLIVEIRA U D, BOER M D, et al.Thick Co-based coating on cast iron by side laser cladding: Analysis of processing conditions and coating properties［J］.Surface & Coatings Technology, 2007, 201(12): 5875-5883.

[7] 朱刚贤, 张安峰, 李涤尘.激光熔覆工艺参数对熔覆层表面平整度的影响［J］.中国激光, 2010, 37(1): 296-301.

ZHU G X, ZHANG A F, LI D C.Effect of laser cladding process parameters on surface smoothness of cladding layer［J］.Chinese Journal of Lasers, 2010, 37(1): 296-301.

[8] KIM J D, PENG Y.Melt pool shape and dilution of laser cladding with wire feeding［J］.Journal of Materials Processing Technology, 2000, 104(3): 284-293.

[9] CHRYSSOLOURIS G, ZANNIS S, TSIRBAS K, et al.An experimental investigation of laser cladding［J］.CIRP Annals-Manufacturing Technology, 2002, 51(1): 145-148.

[10] 常明, 张庆茂, 廖健宏, 等.脉冲Nd:YAG激光熔覆修复塑料模具工艺参数的优化［J］.强激光与粒子束, 2006, 18(9): 1443-1446.

CHANG M, ZHANG Q M, LIAO J H, et al.Optimization of process parameters of pulsed Nd:YAG laser cladding for plastic mold restoration［J］.High Power Laser and Particle Beams, 2006, 18(9): 1443-1446.

[11] 丁周华, 郑启光, 童杏林, 等.基于神经网络的Co基硬质合金激光熔覆工艺优化［J］.光电子·激光, 2004, 15(3): 352-355.

DING Z H, ZHENG Q G, TONG X L, et al.Optimization of laser cladding technology for Co-based hard alloy based on neural network［J］.Journal of Optoelectronics·laser, 2004, 15(3): 352-355.

[12] 飞思科技产品研发中心.神经网络理论与 MATLAB7 实现［M］.北京: 电子工业出版社, 2005.

PHASE ONE TECHNOLOGY R & D CENTER.Neural network theory and MATLAB7 implementation［M］.Beijing: Publishing House of Electronics Industry, 2005.

[13] 周开利, 康耀红.神经网络模型及其 MATLAB 仿真程序设计［M］.北京: 清华大学出版社, 2005.

ZHOU K L, KANG Y H.Neural network model and MATLAB simulation program design［M］.Beijing: Tsinghua University Press, 2005.

[14] MATLAB中文论坛.MATLAB 神经网络 30 个案例分析［M］.北京: 北京航空航天大学出版社, 2010.

MATLAB CHINESE FORUM.MATLAB neural network 30 case studies［M］.Beijing: Beijing University of Aeronautics and Astronautics Press, 2010.

[15] LEARDI R.Genetic algorithms in chemistry［J］.Journal of Chromatography A, 2007, 1158(1): 226-233.

[16] 董晓菲, 张冠伟.基于计算机技术45钢激光熔覆工艺的数值模拟［J］.热加工工艺, 2014, 43(2): 13-16.

DONG X F, ZHANG G W.Numerical simulation of laser cladding process for 45 steel based on computer technology［J］.Thermal Processing, 2014, 43(2): 13-16.

刘干成, 黄博. 基于GA-BP神经网络的镍基合金熔覆涂层形貌预测[J]. 应用激光, 2018, 38(4): 527. Liu Gancheng, Huang Bo. Prediction of Ni-based Alloy Cladding Coatings Topography Based on GA-BP Neural Network[J]. APPLIED LASER, 2018, 38(4): 527.

基于GA-BP神经网络的镍基合金熔覆涂层形貌预测

关于本站 Cookie 的使用提示

全站搜索

基于GA-BP神经网络的镍基合金熔覆涂层形貌预测

相关论文

相关资讯

关于本站 Cookie 的使用提示

全站搜索