在旋转机械振动的故障诊断中,小波包能量算法具有精度高,识别能力强等优点,但传统的小波包能量算法有其自身存在的缺陷,即没有考虑到每个分解频带上的能量参数在时间轴上的分布特征。基于此,提出一种改进的小波包能量算法,研究结果表明,改进的小波包能量算法比传统的小波包能量算法在振动信号分析与故障诊断中更有效。 In the fault diagnosis of rotating machinery vibration, the wavelet packet energy algorithm has the advantages of high precision and strong recognition ability, but the traditional wavelet packet energy algorithm has its own shortcomings, that is, it does not consider the energy parameter in each decomposition frequency band Timeline distribution. Based on this, an improved wavelet packet energy algorithm is proposed. The results show that the improved wavelet packet energy algorithm is more effective than the traditional wavelet packet energy algorithm in vibration signal analysis and fault diagnosis.