复合材料钻削刀具状态集成学习分类技术

发布时间：2018-01-15 22:36

本文关键词：复合材料钻削刀具状态集成学习分类技术　出处：《宇航材料工艺》2015年06期 　论文类型：期刊论文

【摘要】：刀具状态监测关系到工件加工质量,因此实时掌握刀具的磨损状态具有重要的意义。针对单一模式识别分类器的局限性,提出了基于异态集成学习模型的刀具状态监测系统。在该系统中,根据集成学习基分类器选择原则,选取了支持向量机(SVM)、隐马尔科夫模型(HMM)以及径向基神经网络(RBF)这3个单分类器作为基分类器。为了验证监测系统的有效性,进行了复合材料钻削实验,提取了加工过程中的钻削力信号、振动信号的时域特征并利用局部保持法(LPP)进行了特征选择。通过与单分类器和集成学习分类效果的对比,表明了集成学习模型相比于单一基分类器具有更高的分类精度和更好的稳定性。
[Abstract]:Tool condition monitoring is related to the machining quality of the workpiece, so it is very important to grasp the tool wear state in real time, aiming at the limitation of single pattern recognition classifier. A tool condition monitoring system based on heteromorphic integrated learning model is proposed. In this system, support vector machine (SVM) is selected according to the selection principle of integrated learning basis classifier. Three single classifiers, HMMM) and RBF (radial basis function neural network), are used as basis classifiers. In order to verify the effectiveness of the monitoring system, the drilling experiments of composite materials are carried out. The drilling force signal in machining process is extracted, the time domain feature of vibration signal is extracted, and the feature selection is carried out by using local holding method (LPP). The results are compared with that of single classifier and integrated learning classifier. It is shown that the ensemble learning model has higher classification accuracy and better stability than the single base classifier.
【作者单位】：天津大学机械工程学院;
【基金】：国家自然科学基金(51175371) 国家重大专项(2014ZX04012014) 天津市科技支撑计划项目(14ZCZDGX00021)
【分类号】：TG71
【正文快照】： 0引言随着航空航天技术的发展,难加工材料的使用也越来越多。在难加工材料切削过程中刀具磨损比较严重,而刀具的磨损直接关系到表面完整性和加工质量,因此实现刀具磨损状态的监测对提高加工精度至关重要。刀具磨损状态监测方法被分为直接法和间接法[1]。由于直接法具有无法在

【相似文献】