裂纹转子非线性振动特征的谐波小波与分形识别

通过对谐波小波的分析研究,指出了谐波小波对振动信号局部频段分析的优良特性。采用谐波小波对裂纹转子的非线性振动信号在低频段进行了分析。理论分析与实验结果表明:对实际裂纹转子信号,经谐波小波的频段分析后,能够得到通常难以由理论分析与实验结果获得的非整数倍周期分叉的非线性特征频谱。计算了裂纹转子的分形维数,发现裂纹转子的振动信号比理论结果要复杂、可以用多重分形作为判断实际转子是否有裂纹的一个特征。提出了用谐波小波变换后的奇异谱来识别裂纹转子非整数倍周期分叉的非线性特征频谱的方法,并对实际转子信号进行分析,得到了明显的非整数倍周期分叉的奇异谱。

Identification of Non-Liner Vibration Characteristics of Cracked Rotor Using Harmonic Wavelet Analysis and Fractal Theory

The investigation of harmonic wavelet analysis revealed good characteristics in vibration signal analysis on partial frequency domain.Harmonic wavelet analysis was applied to the non-linear vibration signal analysis in the low frequency segment of a cracked rotor.The results of theoretical calculation and experiment show that for a cracked rotor,it is possible to obtain the non-linear character spectrum of non-integral multiple period bifurcations with the help of harmonic wavelet analysis on partial frequency domain.Comparison was made between the calculated and experimental results.It was found that the actual vibration signal of a cracked rotor is more complex than that of calculated.The multi-fractal can be used as an index of whether crack exists in a rotor.The singularity spectrum method obtained from Harmonic wavelet transformation can be regarded as a new identification method and it is put forward for non-liner character spectrums of non-integral multiple period bifurcation identification.Using this new method, the vibration signal of an actual cracked rotor was analyzed.The result is satisfactory.