为了增强轴承内外圈传动振动阶段的故障诊断效果，选择自适应匹配追踪(SAMP)算法构建得到故障诊断模型，完成仿真分析并通过实验进行了验证。研究表明：滚动轴承外圈达到了1 000 Hz的固定频率，同时在固定频率附近形成了1 000±50 Hz边频带；轴承外圈通过频率为40 Hz，在轴承外圈形成了通过频率的2倍与5倍频；对轴承内圈仿真加噪信号谱图通过SAMP算法进行重构，得到轴承内圈通过频率为60 Hz，形成了轴承内圈通过频率的5倍频，轴承内圈通过频率附近形成了60±10 Hz边频带，获得了明显的故障特征；在SAMP重构故障信号谱图中形成了360 Hz的啮合频率与180 Hz的半频，形成了转频的半频10 Hz与3倍频，而啮合频率360 Hz形成了较低的峰值。经SAMP算法处理信号谱图能很好地体现故障特征。

In order to enhance the fault diagnosis effect of bearing inner and outer ring transmissions in vibration stage, the adaptive matching pursuit (SAMP) algorithm was selected to build the fault diagnosis model, the simulation analysis was completed and the experimental verification was carried out. The research results show that the outer ring of the rolling bearing reaches a fixed frequency of 1 000 Hz, and at the same time, a side frequency band of 1 000±50 Hz is formed around the fixed frequency. The passing frequency of the outer ring of the bearing is 40 Hz, which forms 2 times and 5 times of the passing frequency of the bearing outer ring. After the SAMP algorithm is used to reconstruct the simulation and noise signal spectrum of the bearing inner ring, it is found that the passing frequency of the bearing inner ring is 60 Hz, forming 5 times of the passing frequency of the bearing inner ring, and a side frequency band of 60±10 Hz is formed around the passing frequency of the bearing inner ring, obtaining obvious fault features. Experimental verification shows that in the SAMP reconstruction fault signal spectrum, 360 Hz meshing frequency and 180 Hz half frequency are formed, and 10 Hz and triple frequency of half frequency are formed, while 360 Hz meshing frequency forms a low peak value. The signal spectrum processed by SAMP algorithm can reflect the fault characteristics well.

TK221

国家自然科学基金（51675383）；河南省科技攻关项目（182102310706）；南阳市科技攻关项目（KJGG019）