基于反向传播神经网络的压制成型工艺参数优化

Optimization of Processing Parameter for Pressure Molding Based on BP Neural Network

利用反向传播神经网络的可预测性,基于Matlab软件进行压制成型工艺参数优化。以成型温度、成型压力、成型时间及升温速率这四个工艺参数为输入因素,以结合强度、摩擦因数和磨损量这三个性能评价指标为输出参数,建立反向传播神经网络模型,进行训练学习与仿真计算,并进行检验。通过这一反向传播神经网络模型,可以预测不同工艺参数组合下的压制成型制品性能评价指标。通过研究确认,当成型温度为332.32~348.04℃,成型压力为9.39MPa^9.84MPa,成型时间为48.87~51.18min,升温速率为5.86~6.14℃/min时,压制成型的金属塑料自润滑复合材料综合性能最佳。

Based on the predictability of BP neural network, the parameters of pressure molding process were optimized based on Matlab software. Four process parameters i.e. molding temperature, molding pressure, molding time and heating rate were taken as input factors, and three performance evaluation indexes i.e. bonding strength, friction factor and wear extent were taken as output parameters to establish a BP neural network model for training & learning and for simulation & calculation and testing as well. Through this BP neural network model, it is possible to predict the performance evaluation index of pressure molding products under different combinations of process parameters. Through investigation, it is confirmed that when the molding temperature is 332.32~348.04 ℃, the molding pressure is 9.39 MPa^9.84 MPa, the molding time is 48.87~51.18 min, and the heating rate is 5.86~6.14 ℃/min, the metal/plastic self=lubricating composite material made by pressure molding provides the best combination properties.