深孔加工刀具磨损研究与状态识别

发布时间：2018-01-28 19:41

本文关键词： 单齿BTA钻刀具磨损小波分析神经网络　出处：《中北大学》2017年硕士论文　论文类型：学位论文

【摘要】：深孔加工在机械制造业中占据有重要地位,深孔刀具则是深孔加工的核心部件。BTA深孔钻作为深孔加工的常用刀具,最容易出现的问题是刀具磨损,直接影响着加工质量。因此对深孔刀具磨损情况进行研究并能够及时监测到钻头的磨损状态,进行刀具磨损状态的识别具有深远的意义。论文介绍了单齿BTA内排屑深孔钻的结构,将刀具磨损状态分为正常磨损、过度磨损和崩刃三类,并分析了其磨损形式、影响磨损的因素和磨损机理。使用Deform-3D有限元软件对深孔加工过程进行模拟仿真,并根据运行得到的结果分析了加工过程中的温度场分布和刀具磨损情况。论文以BTA内排屑深孔钻削系统为研究对象,在对深孔加工特点以及工况信号进行分析的基础上,建立了以切削功率为监测信号的深孔刀具状态监测系统,并对采集到的功率信号进行了消噪处理和特征提取,最后对特征值进行识别。由于加工环境的复杂性,直接采集的功率信号含有非平稳特性以及包含干扰噪声,必须首先进行处理,因此进行了消噪过程。然后使用小波分析的方法对消噪后的功率信号进行特征提取。利用小波变换原理对功率信号进行多层分解与重构,提取出与刀具磨损状态有较强相关性的高频频带能量,将其作为特征向量。最后建立基于RBF神经网络的深孔加工刀具磨损识别模型,实现特征向量向刀具磨损状态的映射。结果表明,该模型对深孔钻削中刀具磨损状态能够较好的进行识别。
[Abstract]:Deep hole machining occupies an important position in the mechanical manufacturing industry. Deep hole tool is the core part of deep hole machining. BTA deep hole drill is the common tool for deep hole machining. The most common problem is tool wear. It has a direct impact on the machining quality. Therefore, the wear of the deep hole tool can be studied and the wear state of the bit can be monitored in time. It is of great significance to recognize the tool wear state. The structure of single-tooth BTA deep hole drill is introduced in this paper. The tool wear state is divided into normal wear, excessive wear and breakage. The wear form, the factors affecting the wear and the wear mechanism are analyzed. The deep hole machining process is simulated by using Deform-3D finite element software. According to the results obtained from the operation, the temperature field distribution and tool wear in the machining process are analyzed. The paper takes the BTA deep hole drilling system as the research object. On the basis of analyzing the characteristics of deep hole machining and the signal of working condition, a condition monitoring system of deep hole cutting tool is established, which takes cutting power as the monitoring signal. The acquired power signal is de-noised and feature extracted. Finally, the characteristic value is identified. Because of the complexity of the processing environment, the directly collected power signal contains non-stationary characteristics and interference noise. It must be processed first, so the process of de-noising is carried out. Then wavelet analysis is used to extract the features of the de-noised power signal, and the wavelet transform principle is used to decompose and reconstruct the power signal. The high frequency band energy which has strong correlation with the tool wear state is extracted and used as the eigenvector. Finally, the recognition model of tool wear for deep hole machining based on RBF neural network is established. The mapping of feature vector to tool wear state is realized. The results show that the model can recognize the tool wear state in deep hole drilling.
【学位授予单位】：中北大学
【学位级别】：硕士
【学位授予年份】：2017
【分类号】：TG713

【参考文献】