基于小波神经网络预测单个碳纳米管的场增强因子和开启电压

Prediction of Field Enhancement Factor and Onset Voltages of Single Carbon Nanotubes by Wavelet Neural Networks

针对单个碳纳米管的场发射特性参数与几何结构之间的复杂非线性关系,构造了一种多输入多输出的小波神经网络结构来预测单个碳纳米管的场发射增强因子和开启电压,并采用BP算法和自适应的学习速率加快网络训练的收敛速度.网络的训练结果和测试结果表明,预测输出值与实验值之间的误差在2.36%以内,充分说明小波神经网络可以较好地预测碳纳米管的场发射特性.

Based on the complex nonlinear relation between field emission parameters and geometry of carbon nanotubes（ CNTs）, we construct a model to predict the field enhancement factor and the onset voltages by a kind of multi-input-multi-output continual wavelet neural networks（WNN）. BP arithmetic is used to train it and self-adaptive learning rate are also used to accelerate the learning speed.The simulation resuits show that difference between predicted values of WNN and the experimental values is less than 2.36% .WNN can predict the properties of field emission of CNTs very well.