应用BP神经网络预测镍基合金薄壁筒胀形回弹率
Prediction on Rebound Rate of Thin Cylinder of Nickel-based Alloy by BP ANN
李新和， 王艳芬， 杨新泉
点击：3751次 下载：0次
作者单位：中南大学 机电工程学院，湖南 长沙 410083）
中文关键字：神经网络； 薄壁筒； 胀形； 回弹率
英文关键字：artificial neural network (ANN); thin cylinder; bulging; rebound rate
中文摘要：采用BP神经网络方法建立了镍基合金薄壁筒胀形回弹率预测的神经网络模型。模型的输入参数包括胀形温度、胀形压力、摩擦系数、薄壁筒与模具间隙等，模型的输出为薄壁筒中心截面平均回弹率。与传统回归拟合公式相比，该模型具有容错性好、通用性强等优点。该模型可以预测镍基合金薄壁筒在不同胀形工艺参数匹配下的回弹率，也可以用于优化胀形工艺参数。
英文摘要：A neural network model was set up by BP ANN to predict bulging rebound rate of thin cylinder of nickel-based alloy. The inputting parameters of the model include temperature, pressure, friction coefficient and clearance between thin cylinder and die, the outputting parameters are the average rebound rate of center section in the thin cylinder. Compared with traditional regression fitting formula, the model has good fault tolerance, versatility and other advantages. The model can predict the bulging rebound rate of thin tube of nickel-based alloy at different process parameter, it can also be used to optimize the process parameters of bulging.