论文部分内容阅读
以油箱端盖作为分析对象,借助DYNAFORM仿真软件,对油箱端盖的拉深成形过程进行数值模拟,并通过拉深成形试验验证可知,板料最大减薄率与最大增厚率的试验值与模拟值之间的相对误差分别为9.26%与8.32%,验证了有限元模型的正确性。结合正交试验,进行有限元仿真试验的设计,基于BP人工神经网络,对板料的成形质量进行仿真预测。选择冲压速度、模具间隙以及压边力作为输入层,将板料成形的最大减薄率作为输出层,建立了3-11-1的3层BP人工神经网络。通过BP人工神经网络的训练与测试得知:BP人工神经网络仿真预测值与数值模拟值之间的相对误差为2.15%,验证了BP人工神经网络应用于油箱端盖拉深成形质量仿真预测的正确性。
Taking the fuel tank end cap as the analysis object and the DYNAFORM simulation software, the numerical simulation of the deep drawing process of the oil tank end cap was carried out. Through the verification of the deep drawing test, it can be seen that the maximum and the maximum thickening rates The relative errors between the simulated values were 9.26% and 8.32%, respectively, verifying the correctness of the finite element model. Combined with orthogonal test, the design of finite element simulation test was carried out. Based on BP artificial neural network, the forming quality of sheet metal was simulated and predicted. Select the stamping speed, die clearance and blank holder force as the input layer, the sheet forming the maximum reduction rate as the output layer, the establishment of 3-11-1 3-layer BP artificial neural network. Through BP neural network training and testing, we know that the relative error between BP artificial neural network simulation prediction value and numerical simulation value is 2.15%, which verifies that BP artificial neural network is used to predict the quality of oil tank end cap forming Correctness