Prediction Model and Experimental Verification of Residual Stress in Micro-forging of TC4 Titanium Alloy
In order to predict the residual stress field introduced by the micro-forging process of TC4,a combined model based on convolutional neural network(CNN)and short and long time memory network(LSTM)is proposed in this paper.The CNN module extracts high-dimensional features from the input data,and then the LSTM module carries out serialization modeling.The TC4 micro-forging finite element model is established and the accuracy of the finite element model is verified by the micro-forging experiment.The data set is established based on the finite element simulation and the model is trained.The results show that the performance indices of the model are superior to the LSTM model.The prediction performance of the model is verified by the micro-forging experiment,which shows that the model has certain validity and practicability.
NETL
NSTL
万方数据
