为了正确提取磁悬浮轴承跌落过程中发生反向涡动时轴轨迹的时频特征,本文引入了广义线性调频小波变换(GLCT).为分析不同时频方法在信号时频特性分析的优缺点,分别采用了短时傅里叶变换、Wigner-Ville分布、GLCT和Hilbert transform对磁悬浮轴承跌落过程中反向涡动情况下轴系Y轴(竖直)方向上的位移信号进行了处理.结果显示,短时傅里叶变换因为分析窗不具备瞬时频率调频率的自适应性,时频聚焦性不理想,瞬时频率的计算精度易受其影响;Wigner-Ville分布中交叉项会混淆真实时频分布,也会带来计算误差;而Hilbert transform过分依赖信号光滑性,噪声影响较大;GLCT具备理想的时频聚焦性,易于获得正确的瞬时频率.这有利于后续重新悬浮控制程序的开展.%The general linear chirple transform(GLCT) is introduced to extract the time-frequency characteristic of the rotor orbit during full rub in active magnetic bearings. In order to analyze the advantages and disadvantages of different time-frequency methods in the analysis of time-frequency characteristics of the signal,the short-time Fourier transform,Wigner-Ville distribution,GLCT and Hilbert-Huang transform are used to process the rotor displacement signal in Y(vertical) direction during full rub. The results show that the short-time Fourier transform is not ideal in time-frequency focusing because the analysis window does not have the adaptive frequency modulation of the instantaneous frequency. The accuracy of the instantaneous frequency is susceptible to its influence. The cross term in the Wigner-Ville distribution confuses the true time-frequency distribution and brings the calculation error. Hilbert-Huang transform is too dependent on the smoothness of signal,noise will influence the results significantly. GLCT has the ideal time-frequency focusing,easy to obtain the correct instantaneous frequency. This is conducive to the subsequent re-levitation control program.
展开▼
