摘要
数控机床刀具磨损监测对于提高刀具利用率,减小因刀具磨损而造成的损失具有重要意义。基于对电流信号特点的分析和小波包分解技术对信号特征量提取的优势,提出一种通过监测机床主轴电机电流特征量变化来监测刀具磨损状态的方法。该方法利用db8小波基对电流信号进行4层小波包分解,将分解后各频带上的均值与方差作为特征量。建立从新刀到刀具磨损状态下特征量随刀具切削时间的变化关系,根据特征量的变化即可判别刀具磨损状态。试验结果验证了该方法在刀具磨损监测中的可行性。
CNC tool wear monitoring has great importance in improving the utilization of machine tools and in reducing the economic losses due to the tool wear. Based on the characteristic analysis of current signals and the advantages of wavelet packets decomposition theory in the signal feature extraction, a method to judge the tool wear state by means of the current signals is proposed. The method uses db8 wavelet packet to decompose current signal into 4 levels, taking the decomposed signal mean and variance in frequency bands as the features, and then use these features to establish the relationship between the feature and the cutting time from new tool condition until tool wear condition. So that the tool wear condition can be distinguished according to the variance contribution of feature. The usability of this method is verified by the test results.
出处
《心智与计算》
2010年第4期258-264,共7页
Mind and Computation
关键词
刀具磨损
电流信号
小波包分析
特征提取
tool wear
current signal
wavelet packet analysis
feature extraction