论文部分内容阅读
针对信号经验模态分解(EMD)过程中存在波形混叠现象,提出一种基于聚合经验模态分解(EEMD)和Hilbert边际谱相结合的方法对齿轮箱故障进行故障诊断。首先使用小波阈值分析对背景噪声较大的齿轮箱振动信号进行预处理,提高EEMD分解的精确度;其次对预处理信号进行分解,得到IMF分量,对比正常信号与故障信号的区别;最后对2种工况信号进行Hilbert变换并计算得到边际谱,确定故障信号的故障频率。研究表明该方法在避免EMD分解带来的模态混叠现象方面具有可行性,能提高齿轮箱故障诊断的准确率。
In order to deal with the phenomenon of waveform aliasing in the EMD process, a fault diagnosis method based on the combination of empirical mode decomposition (EEMD) and Hilbert marginal spectrum is proposed. Firstly, wavelet threshold analysis is used to preprocess gearbox vibration signals with high background noise to improve the accuracy of EEMD decomposition. Secondly, the preprocessed signals are decomposed to obtain IMF components, and the difference between normal signals and fault signals is compared. Finally, Hilbert transform is performed on the working condition signals and the marginal spectrum is calculated to determine the fault frequency of the fault signal. Research shows that this method is feasible in avoiding modal aliasing caused by EMD decomposition and can improve the accuracy of gearbox fault diagnosis.