基于EMD和小波分解的颤振信号特征提取方法比较研究

Comparative Study on Chatter Signal Feature Extraction Methods Based on EMD and Wavelet Transform

以数控动梁龙门导轨磨床KD4020X16为试验对象,对不同磨削参数情况下磨床的不同振动状态进行试验测试。分别应用经验模态分解和小波变换对磨床的颤振信号进行颤振特征提取,以相关系数和方差为依据分析和比较这两种方法在非线性时序信号处理中的优劣。分析结果表明,在磨床的磨削颤振信号特征量提取方面,基于经验模态分解的提取方法具有更好的准确性和自适应性。

With numerical control and moving beam-type planer guideway grinder KD4020X16as test object,this paper tests different vibrational states of grinder under different grinding parameters,respectively uses empirical mode decomposition and wavelet transform for chatter feature extraction of chatter signal of grinder and analyzes and compares advantages and disadvantages of both methods in non-linear sequence signal processing based on correlation coefficient and variance.The analysis result shows that the extraction method based on empirical mode decomposition has higher accuracy and adaptivity in terms of the extraction of characteristic quantity of grinding chatter signal of grinder.