用人工神经网络预测电铸自支撑金刚石-镍复合膜沉积结果

Prediction Result of Free-Standing Diamond Grains-Nickel Composite Film by Electrotyping Using an Artificial Neural Network

用人工神经网络预测了电铸自支撑金刚石-镍复合膜中金刚石颗粒的含量、复合膜的厚度和表面微观形貌。结果表明,当阴极电流密度小于1.0A/dm2时,复合膜的表面均匀,无镍瘤;复合膜的沉积速率约为14μm/h。其预测的沉积结果与实际样品测量值接近,相对误差小于9.9%。人工神经网络能够充分体现电镀工艺参数与沉积结果之间的非线性关系和隐含关系,训练精度较高,具有较高的预测能力。

An artificial neural network (ANN) was used to predict diamond grains content, thickness and micro morphology of free-standing diamond grains-nickel composite film by electrotyping. It is found that the predicted results are near the measured values, and the relative error is less than 9.9%. The dense and uniform composite films without nickel burls are obtained when the current 2 densities are less than 1.0 A/dm(2); and the deposited velocity of the film is about 14 mu m/h. ANN is useful to map the non-linear relationship among electrodeposited parameters and the results, and this technique is more accurate.