基于BP神经网络模型的Ni-SiC纳米镀层耐磨性能预测研究  

Study on the wear resistance prediction of Ni-SiC nanocoating based on a BP neural network model

作　　者：李心源 马春阳[2] 赵旭东 LI Xinyuan;MA Chunyang;ZHAO Xudong(Information Center, Daqing Normal University, Daqing 163712 China;College of Mechanical Science and Engineering, Northeast Petroleum University, Daqing 163318, China)

机构地区：[1]大庆师范学院信息中心,黑龙江大庆163712 [2]东北石油大学机械科学与工程学院,黑龙江大庆163318

出　　处：《功能材料》2020年第1期1126-1130,共5页Journal of Functional Materials

基　　金：国家自然科学基金资助项目(51974089);黑龙江省自然科学基金资助项目(LC2018020)

摘　　要：采用神经网络技术,构建结构为3×8×1型的BP神经网络模型,并利用该模型对超声电沉积Ni-SiC纳米镀层的耐磨性能进行预测。通过磨损试验测试并研究Ni-SiC纳米镀层的耐磨性能,利用扫描电镜(SEM)、原子力显微镜(AFM)和X射线衍射(XRD)观察不同参数下Ni-SiC纳米镀层的组织结构及成分。结果表明,在BP神经网络模型的隐含层数和神经元数分别为1和8时,该BP神经网络模型的均方根误差最小,其最小值为1.24%。该BP神经网络模型的预测值与实验值相差不大,其最大误差为1.51%。当采用SiC粒子浓度8 g/L、电流密度2 A/dm^2、温度40℃时,SiC粒子均匀分布于Ni-SiC纳米镀层中,且镀层镍晶粒显著细化,其镍晶粒的衍射峰变宽、变矮。The BP neural network model with a structure of 3×8×1 was established by using artificial neural network technology.The wear resistance of the coating prepared by ultrasonic electrodeposition was predicted by this model.The wear resistance of Ni-SiC nanocoating was tested by a wear test and the microstructure and component of the coatings prepared at different parameters were observed via scanning electron microscope(SEM),atomic force microscope(AFM)and X-ray diffraction(XRD).The results showed that when the number of hidden layers was 1 and the number of neurons was 8,the root mean square error of the BP neural network model was the smallest,and the minimum value was only 1.24%.The predicted value of the BP neural network model was not much different from the experimental value,and the maximum error was 1.51%.When the concentration of SiC particles was 8 g/L,the current density was 2 A/dm^2,and the temperature was 40℃,the SiC particles were uniformly distributed in the Ni-SiC nanocoating.The nickel grains of the coating were obviously refined,and their diffraction peak became wider and shorter.

关 键 词：超声电沉积 BP神经网络模型 Ni-SiC纳米镀层 

分 类 号：TG174.4[金属学及工艺—金属表面处理]