基于神经网络的镍基涂层转接工艺分析与模拟  

Analysis and Simulation of Coating Transfer Process Based on Neural Network

作　　者：朱昱[1] 李小武[1] 孙书刚 倪红军[1] Zhu Yu;Li Xiaowu;Sun Shugang;Ni Hongjun(School of Mechanical Engineering,Nantong University;Nantong Gaoxin Wear-resistant Polytron Technologies Inc.)

机构地区：[1]南通大学机械工程学院 [2]南通高欣耐磨科技股份有限公司

出　　处：《特种铸造及有色合金》2018年第12期1285-1288,共4页Special Casting & Nonferrous Alloys

基　　金：江苏省重点研发计划资助项目(BE2016107)

摘　　要：为了探究涂层转接工艺及配方对转接涂层性能的影响,利用神经网络模拟涂层配方、涂层转接工艺与复合涂层性能的关系模型的有效性,并用测试数据验证模型。结果表明,转接涂层硬度、磨损量、剪切强度的模拟值与试验值的相对误差最大值分别为6.19%、9.52%和6.45%。采用BP神经网络模型能够很好地模拟涂层转接工艺、配方与转接涂层硬度、耐磨性及剪切强度间的内在联系,验证了模拟结果的准确性。In order to understand the influence of coating transfer technology and formula on the performance of the transfer coating,the neural network was used to simulate the coating formula,the relationship model of the coating transfer process and the performance of the composite coating.Besides,the model was verified by the test data.The results reveal that the maximum relative error of the coating hardness,the wear rate,the shear strength reach 6.19%,9.52% and 6.45%,respectively.The BP neural network model can well simulate the internal relationship between coating transfer process,formulation and hardness,wear resistance and shear strength of transfer coating,and it is effective to the analysis and simulation of the coating transfer process.

关 键 词：涂层转接工艺 复合涂层 神经网络 工艺模拟 

分 类 号：TG335.86[金属学及工艺—金属压力加工] TG146.15[一般工业技术—材料科学与工程]