基于进化BP神经网络的磨削温度预测研究

Research on Grinding Temperature Prediction Based on Evolutionary BP Neural Network

磨削过程中,过高的磨削温度会对零件产生热损伤等负面影响,为了能够掌握并控制磨削温度,对磨削温度场进行了仿真和预测研究。首先,利用有限元法对磨削温度进行仿真;随后,分别采用BP和遗传算法优化BP神经网络进行仿真结果预测。最后,将两种神经网络预测结果分别与仿真结果进行比较。结果显示,遗传算法优化BP神经网络性能更优,预测值更接近仿真值。

During the grinding process,too high grinding temperature will have a negative impact on the thermal damage of the part.In order to grasp and control the grinding temperature,the grinding temperature field is simulated and predicted.Firstly,the finite element method is used to simulate the grinding temperature.Subsequently,BP and genetic algorithm are used to optimize the BP neural network for simulation prediction.Finally,the two neural network prediction results are compared with the simulation results.The results show that the genetic algorithm optimizes BP neural network performance better,and the predicted value is clos⁃er to the simulation value.