基于小波包及Hilbert-Huang变换的数控铣削颤振诊断技术

发布时间：2018-07-08 08:00

本文选题：颤振 + 诊断　；参考：《计算机集成制造系统》2015年01期

【摘要】：为了提高产品加工质量,根据试验测得铣削系统颤振稳定域,制定并采集数控铣削振动信号,以保证采集信号的准确性;融合小波包变换与希尔伯特黄变换,从能量频域分布与幅值概率统计分布两方面提取信号特征值,其中小波包降噪作为信号前置处理能有效降低环境噪声干扰的影响,提高经验模式分解的精度;建立基于模糊支持向量机的颤振诊断模型,将振动信号分为平稳铣削信号、微弱颤振铣削信号、颤振铣削信号及刀具磨损铣削信号。实验结果表明,该模型具有良好的铣削振动信号辨识与诊断能力,预测准确率达97.3%,为数控铣削加工振动信号的准确辨识与诊断提供了一种新方法。
[Abstract]:In order to improve the machining quality of the product, the chatter stability region of the milling system is measured according to the experiment, and the vibration signal of NC milling is formulated and collected to ensure the accuracy of the acquisition signal, and the wavelet packet transform and Hilbert yellow transform are combined. The signal eigenvalues are extracted from the energy frequency domain distribution and the amplitude probability statistical distribution. The wavelet packet denoising as a signal preprocessing can effectively reduce the influence of environmental noise and improve the precision of empirical mode decomposition. The flutter diagnosis model based on fuzzy support vector machine is established. The vibration signal is divided into stationary milling signal, weak chatter milling signal, flutter milling signal and tool wear milling signal. The experimental results show that the model has a good capability of milling vibration signal identification and diagnosis, and the prediction accuracy is 97.3. it provides a new method for accurate identification and diagnosis of NC milling vibration signal.
【作者单位】：北京航空航天大学机械工程及自动化学院;北京航空航天大学北京市高效绿色数控加工工艺及装备工程技术研究中心;
【基金】：国家自然科学基金重大资助项目(11290144) 国防基础科研计划资助项目(A2120110002)~~
【分类号】：TG547

【共引文献】