应用小波包分解能对信号高、低频部分局部细化并保留原信号的时域特征,因而具有良好的时频局部化特性,对非平稳信号进行有效识别的技术,从采集到的有叶片裂纹时的振动信号,通过选取适当的频段进行信号重构,提取叶片裂纹故障的特征,从而实现转子叶片裂纹故障诊断。 The application of wavelet packet decomposition can refine the high frequency part and low frequency part of the signal and preserve the time-domain features of the original signal. Therefore, it has good time-frequency localization features and effectively identifies non-stationary signals. When the cracked blade is collected, The vibration signal of the rotor blade can be diagnosed by selecting the appropriate frequency band for signal reconstruction and extracting the characteristics of the blade crack fault.