充分利用小波包(WP)的时频局部化分析能力和最大熵谱估计(MESE)的频谱细化优点,提出了一种新的谱分析方法——小波包-最大熵谱估计(WP-MESE)。该方法能对振动信号进行多分辨分解,提取故障特征信息,进而进行精确的频谱分析;并运用此法有效地提取了水轮机轴系动态特性信息。实践证明该法是提取微弱故障信息并进行早期诊断的有效方法。 Taking full advantage of the time-frequency localization analysis ability of wavelet packet (WP) and the spectral refinement of maximum entropy spectrum estimation (MESE), a new spectral analysis method-wavelet packet-maximum entropy spectrum estimation (WP-MESE ). The method can decompose the vibration signal by multi-resolution and extract the fault feature information for accurate spectrum analysis. The method can effectively extract the dynamic characteristic information of the turbine shaft. Practice has proved that the method is to extract weak fault information and effective early diagnosis method.