基于动态反馈神经网络模型的数控机床热误差实时预报补偿  

Real-time predictive compensation for thermal error of NC machine tool based on dynamic feedback neural network model

作　　者：唐治[1] 闫开印[1] 田怀文[1] 

机构地区：[1]西南交通大学机械工程学院,成都610031

出　　处：《机械设计与制造》2007年第8期119-121,共3页Machinery Design & Manufacture

摘　　要：分析了数控机床热误差对制造加工精度的影响及影响数控机床热误差的关键因素,对采用神经网络模型预报数控机床热误差进行了深入分析,提出了具有反馈输入环节的动态神经网络模型。讨论了建立模型的关键技术问题,并对基于该模型的数控机床热误差的智能预报进行了计算机仿真分析。The effect of machining accuracy caused by thermal error of NC machine tool and the key factors which lead to thermal error are analyzed. The thermal error of NC machine tool predicted by neural network is lucubrated in this paper. A dynamic feedback neural network model is presented. The establishment of this model, some key technical problems in this system are discussed. Finally, the model was tested by a set of experimental data The results testify it is efficient that this model predicts and compensates NC machine tool thermal error.

关 键 词：动态反馈神经网络 实时预报补偿 数控机床 热误差 

分 类 号：TH161[机械工程—机械制造及自动化] TG659[金属学及工艺—金属切削加工及机床]