用BP神经网络模型预测Ni-Al2O3复合涂层Al2O3粒子复合量研究

运用人工神经网络技术建立一个结构为4×9×1型的BP神经网络模型,用该模型对Ni-Al2O3复合涂层中Al2O3复合量进行预测研究,并用XRD衍射仪和原子力显微镜(AFM)对Ni-Al2O3复合涂层的Al2O3复合量和立体形貌进行分析.结果表明,当隐含层数为9个时,BP神经网络的均方根误差最小(1.13;),BP神经网络的拟合相似度R=0.99937,这表明BP神经网络模型能够较好的预测涂层Al2O3复合量.当占空比60;、阴极电流密度4 A/dm2、pH值4、镀液温度55 ℃时,Ni-Al2O3复合涂层结构密实,结晶细致.

Study on the Al2O3 Contents in Ni-Al2O3 Composite Coatings Predicted by Using BP Neural Network Model

A 4×9×1 type of BP neural network model was set up to predict the Al2O3 contents in Ni-Al2O3 composite coatings by using the artificial neural network technology.The content and 3D surface pattern were analyzed by using XRD diffraction and atomic force microscopy (AFM).The results show that when the number of hidden layers is 9, the minimum root mean square error is 1.13%, the fitting similarity R is about 0.99937, which indicates that the BP neural network model can accurately predict the Al2O3 contents.When the duty ratio is 60%, the cathode current density of 4 A/dm2, pH=4, the bath temperature of 55 ℃, Ni-Al2O3 composite coating has a dense structure, and the crystalline is fine.