基于人工神經網絡的熱軋碳鋼變形抗力預報

Prediction of Flow Stress of Carbon Steel in Hot Strip Rolling by Artifcial Neural Network

摘要: 以恒應變速率凸輪壓縮試驗機得到的實驗數據為基礎,采用人工神經網絡的方法建立了碳鋼變形抗力與應變、應變速率及溫度對應關系的預測模型,與多元非線性回歸模型比較,神經網絡模型具有較高的預測精度.

Abstract: On the basis of the data obtained on Cam Plastometer with constant strain rate, the predicting model for the relation between flow stress and deformation strain, strain rate and temperature of carbon steel has been developed by Artificial Neural Network method. Compared with nonlinear regression method, the neural network model gives better results.