论文部分内容阅读
以Gleeble-1500热模拟试验机获得的Ti40钛合金压缩试验数据为基础,应用人工神经网络对数据进行训练和预测,建立该合金的高温流动应力与应变、应变速率和温度对应关系的预测模型,其中,应变、应变速率(对数形式)和变形温度作为模型的输入参数,流动应力作为模型的输出参数。结果发现,运用BP反向传播算法进行训练的神经网络模型具有良好的预测功能,其预测值与实验测量值基本吻合。同时,采用神经网络模型预测的数据构造Ti40合金的加工图,其安全区和失稳区的范围与实测数据获得的加工图基本相符,并对各自区域的相应组织状态进行金相观察。
On the basis of the compressive test data of Ti40 titanium alloy obtained by Gleeble-1500 thermal simulation testing machine, the data were trained and predicted by using artificial neural network. The predictive model of the corresponding relationship between high temperature flow stress and strain, strain rate and temperature was established. Among them, strain, strain rate (logarithmic form) and deformation temperature are used as input parameters of the model, and flow stress is used as the output parameter of the model. The results show that the BP neural network model trained by the BP back propagation algorithm has a good prediction function, and the predicted value is in good agreement with the experimental measurement value. At the same time, using the data predicted by the neural network model to construct the processing diagram of the Ti40 alloy, the range of the safe zone and the destabilizing zone is basically in accordance with the processing diagram obtained from the measured data, and the corresponding microstructure of the respective zone is observed metallographically.