基于BP神经网络的立铣加工振动快速预估

Rapid Estimation of Vibration of Peripheral Milling Process Based on BP Neural Network

结合研制的立铣加工过程虚拟仿真系统和实验测量铣削力信号,训练并建立优化的1-20-1型BP神经网络模型,快速实现铣削加工过程刀具-工件系统振动状态的预估.对比神经网络模型预估的振动结果与实验测量振动信号可以看出,二者数据吻合较好,表明铣削虚拟仿真系统与神经网络技术的结合能够高效低耗地用于不同铣削加工条件下铣削振动状态的快速预估和加工过程监测.

Combined with a virtual simulation system of peripheral milling process and cutting force signal obtained in cutting trials,an optimal BP neural network model with 1 -20 - 1 structure that can be used rapidly to predict vibration in peripheral milling process is trained and established. The estimation results of vibration displacement obtained from the BP neural network model have a good agreement with the experimental results,which reveals that neural network technique combined with the virtual simulation system can be used effectively to predict and monitor vibration in peripheral milling process under different cutting parameters.