基于BP神经网络的Tb_(0.3)Dy_(0.7)Fe_(1.95)磁致伸缩性能研究

Magnetostriction Properties of Tb_(0.3)Dy_(0.7)Fe_(1.95) Based on Artificial Neural Network

基于BP神经网络算法和理论,研究了Tb0.5Dy0.7Fe1.95合金的磁致伸缩性能,建立了影响磁致伸缩性能参数与伸缩性的预测神经网络模型。结果表明,随着添加元素量的增加,磁致伸缩性在降低;随着磁场强度的增强,磁致伸缩性也随着增强;建立的神经网络模型的预测结果能与实验结果很好对应,误差很小。

Based on BP neural network theory and algorithms, the magnetostriction properties of Tb0.3Dy0.7Fe1.95 were studied. An artificial neural network model for the impact parameters of magnetostriction and the flexibility was built. The experiment results show that with the increase of Ti elements, the magnetostriction reduces. With the enhancement of the magnetic field strength, the magnetostriction enhances. The predicted results correspond well with the experimental results, and the error is very small.