本文从啮合振动信号的分析出发,对齿轮疲劳的早期诊断技术进行了研究。针对齿轮故障所表现的特征,提出了根据峰态系数、幅值概率密度以及峰值谱对齿轮疲劳进行有效地诊断。通过对实际齿轮的试验与跟踪分析,在时域、频域以及幅值域中进行了具体地研究,得到了关于齿轮疲劳诊断的若干有意义的结果。 Based on the analysis of meshing vibration signals, this paper studies the early diagnosis of gear fatigue. According to the characteristics of gear failure, it is proposed to diagnose gear fatigue effectively based on kurtosis coefficient, amplitude probability density and peak spectrum. Through the experiment and tracking analysis of the actual gear, the concrete study has been carried out in the time domain, the frequency domain and the amplitude domain, and some meaningful results about gear fatigue diagnosis have been obtained.