基于残差比阈值优化冲击衰减字典的轴承故障特征提取研究

Research on bearing fault feature extraction based on residue ratiothreshold optimized impact attenuation dictionary

机械设备工作时引起的振动信号蕴藏着丰富的工作状态信息。轴承是机械设备中重要的部件,其工作状态严重影响着设备的工作性能。根据信号特征张成与信号匹配度较高的冲击衰减字典,通过对比几种不同的稀疏表示迭代停止条件的优劣后,选用残差比阈值作为迭代停止条件,并结合稀疏表示中的MP算法进行轴承故障特征的提取研究。分别验证了轴承故障仿真信号、加工故障信号以及加速疲劳失效轴承故障信号,对比传统去噪方法,结果表明其能够在强背景噪声中较为准确的重构出故障特征信号,极大地提高了恶劣工况下故障特征提取的成功率与准确性,同时显示该方法在轴承早期故障信号中能准确可靠地提取出故障频率。

Vibration signals caused by the operating machinery contains a wealth of information on the state of work. Bearing is an important part of mechanical equipment;its working condition seriously affects the working performance of the equipment. In this paper, high signal matched impact attenuation dictionary was constructed according to the signal characteristics. By comparing the advantages and disadvantages of several different iteration stop conditions in sparse representations, the residual ratio threshold was chosen as the iteration stop condition, and the bearing fault feature was extracted combining with the MP algorithm of sparse representation. Then the method is used to verify the bearing fault simulation signal, the machining fault signal and the acceleration fatigue failure bearing fault signal. Compared with the traditional denoising method, the proposed method can reconstruct the fault feature signal more accurately in strong background noise, and improves the success rate and accuracy of fault feature extraction under harsh conditions. Therefore, it can accurately and reliably extract the fault frequency in the early bearing fault signal.