变转速下滚动轴承阶比峭度图法故障特征提取
COT-kurtogram Based Fault Features Extraction for Rolling Element Bearing Under Varying Speed

结合传统阶比分析和峭度图算法的优势，利用计算阶比跟踪方法将时域非平稳信号转换为角域平稳信号，并利用峭度指标准确表征滚动轴承振动信号中的故障瞬态冲击大小，提出了阶比峭度图算法。仿真故障信号及实测滚动轴承外圈故障信号分析结果表明，阶比峭度图算法能够有效识别阶比域内的最优包络解调频带参数，显著提高了变转速工况下滚动轴承故障特征提取的准确性。
This paper converts a time domain non-stationary signal to an angle domain stationary signal using a computed order tracking(COT) method, which combines the advantages of both the traditional order analysis algorithm and kurtogram algorithm. The transient impact force of the fault vibration signal from the rolling bearing was valued by the kurtosis index, and the COT-kurtogram algorithm was thus proposed. Both the simulation and practical results demonstrate that the COT-kurtogram method can effectively identify the envelope demodulation bandwidth parameter in the order-domain, and significantly improve the accuracy of the fault feature extraction of rolling bearing in varying speed working conditions.