摘要
基于切削力分量测量信号,提出了用于端面铣削的刀具状态监测(TCM)的3层BPNN网络系统,用于估计铣削过程中的刀具磨损(Vb)和表面粗糙度(Ra)。利用切削力数据构建了6×10×2结构的神径网络的训练样本,并对其性能进行了评价。建立了刀具磨损和表面粗糙度与有关的切削参数关系。试验结果表明模型输出与直接测量的刀具磨损和表面粗糙度的值非常接近,证明了该方法是可行的。
A three-layered BPNN system based on the measurement of cutting force components for tool condition monitoring （TCM） in the face milling was advanced, and it can be used to estimate flank wear （Vb） and cutting surface roughness （Ra）. The training swatch of neural network configuration as 6 × 10 × 2 was established by cutting force data and the performance was evaluated. The relationship between cutting parameters with Vb and Ra was set up. The testing result show close matching between the model output and Vb and Ra of directly measurement, and also show that the method has significant realistic meaning to tool online monitoring and to advancing cutting quality.
出处
《机床与液压》
北大核心
2010年第5期12-16,11,共6页
Machine Tool & Hydraulics
基金
江苏省教育厅自然科学基金项目(07JKD460075)
关键词
切削力
神经网络
刀具磨损
表面粗糙度
监测
Cutting force
Neural network
Flank wear
Surface roughness
Monitoring
