基于人工神经网络BTA钻削时表面粗糙度的预测  

Surface roughness prediction in BTA drilling based on artificial neural network

作　　者：高腾[1] 苗鸿宾[1] 江敏[1] 

机构地区：[1]中北大学机械工程与自动化学院,山西太原030051

出　　处：《机械设计》2014年第4期4-10,共7页Journal of Machine Design

基　　金：国家自然科学基金资助项目(51175482);山西省国际合作资助项目(2012081030)

摘　　要：为解决深孔加工中表面粗糙度在线检测困难这一问题,提出一种基于BP神经网络的表面粗糙度在线辨识方法,并以BTA钻削为例,建立表面粗糙度BP神经网络在线辨识模型,并将其引入钻削加工领域。该模型能方便地预测钻削加工参数对加工表面粗糙度的影响,有助于准确认识已加工表面质量随切削参数的变化规律,为切削参数的优选和表面粗糙度的控制提供了依据。实验和仿真结果表明,基于BP神经网络模型能够很好地预测表面粗糙度,对提高加工表面粗糙度具有一定的指导意义。In order to conquer the difficulty of on-fine surface roughness measuring, the surface roughness identification method based on BP networks is put forward. As an example, the identifi- cation model of BTA drilling is built and introduced into the field of drilling. The model conveniently predicts the effects of drillingparameters on surface roughness of machined surface, which con- tributes to accurately understand the variation law of quality of ma- chined surface following drilling parameters and provides the foun- dation for properly selecting cutting parameters and controlling surface quality. The simulation and experimental results show that BP neural network can well predict the surface roughness and have a certain guiding significance to improve the surface roughness.

关 键 词：BP神经网络 BTA钻削 表面粗糙度 过程监控 

分 类 号：TH161[机械工程—机械制造及自动化]