基于BP神经网络的Mg-Y-Nd合金热压缩变形流变应力研究

Flow Stress of Mg-Y-Nd Alloy During Hot Compression Deformation Based on BP Neural Network

通过对Mg-Y-Nd-Gd-Zr镁合金进行热压缩变形实验,测得了其在不同流变数率、温度和应变条件下的流变应力。使用BP神经网络的相关算法和理论建立BP神经网络模型,并用采集到的数据对其进行训练,用建立的BP神经网络模型对实验结果进行预测。结果表明,建立的BP神经网络预测精度很高,误差在5%以内,能够很好地反映实验条件和实验结果的相关规律。

The flow stress of Mg-Y-Nd-Gd-Zr alloy was obtained by hot compression deformation experiments under different flow rates, temperatures and strains. Using BP neural network algorithms and theory, a BP neural network model was established and was trained with the collected data. The experimental results were predicted with the BP neural network model. The results show that the BP neural network model is with a high prediction accuracy, the error is within 5%, and can well reflect the related laws between experiment conditions and experimental results.