摘要
根据热缩加长刀杆与刀具配合精加工与半精加工的特点,利用反向传播神经网络(BPNN)建立高速加工热缩加长刀杆与刀具配合的铣削力模型。模型除了考虑6个主要影响铣削力的加工条件外,还将时间参量引入输入向量,实现了三向铣削力的瞬态预测。通过大量的加工实验获得网络所需的训练和检验样本,并通过编制Matlab程序实现了网络性能评价和网络参数优化。检验结果表明,铣削力预测结果与实际测量结果之间具有很好的一致性,三向分力的平均预测误差均小于0.18,在预测效率和精度上均优于通常所用的解析模型,并具有很好的扩展性能。
Current research focuses on developing a milling force model according to the characteristics of the matching of lengthened shrink-fit holder and cutting tool using back propagation neural network(BPNN).Time parameter is taken as a factor of the input vector besides six processing conditions which mainly influence the milling force,and then the forecasting of 3D transient milling forces are achieved.In order to get training and testing samples,a lot of milling experiments are performed and a Matlab program is designed to evaluate and optimize the network.The test experiments show that the forecasting results are in good agreement with the experimental results and the errors of 3D force components are less than 0.18.In addition to improved performance,the BPNN model has higher efficiency and higher accuracy than the traditional analytical model.
出处
《机械科学与技术》
CSCD
北大核心
2010年第4期504-508,共5页
Mechanical Science and Technology for Aerospace Engineering
基金
国家自然科学基金项目(50605008)
湘潭大学博士启动基金项目(08QDZ41)
广西制造系统与先进制造技术重点实验室开放课题基金项目资助
关键词
铣削力
神经网络
高速加工
热缩加长刀杆
milling force
artificial neural network
high speed milling
lengthened shrink-fit holder