基于时频脊线和阶次分析的转子故障诊断

为解决采用在转子升降速过程中产生的非平稳信号难以进行故障诊断的问题,提出一种基于2维时频脊线和阶次分析的转子故障诊断方法。采集转子升降速信号,采用2维时频分布的峰值脊线提取法获得信号脊线特征,结合脊线特征与等角度重采样技术依次获得信号角度域、角-阶域和阶次域图像,将信号阶次域内的特征参数作为故障敏感特征,输入人工神经网络诊断模型,对转子信号的故障类型进行分类。利用实测信号验证所提方法的实际应用效果,并与传统特征提取法的结果进行对比。结果表明:阶次分析方法的测试准确率约为99.8%,标准差小于0.09%,均优于传统特征提取法。基于时频脊线和阶次分析的转子故障诊断方法具有更高的诊断准确率,在非平稳信号特征提取过程中具有很好的可行性和准确性。

Rotor Fault Diagnosis Based on Time-frequency Ridge and Order Analysis

In order to solve the problem that it was difficult to diagnose the rotor fault by using the nonstationary signals generated in the process of rotor speed fluctuation,a rotor fault diagnosis method based on 2-D time-frequency ridge and order analysis is proposed.Collect updown speed rotor signal.2-D time-frequency distribution peak ridge extraction method is used to obtain signal ridge features.The angle domain,angle-order domain and order domain images of signals are obtained by combining ridge feature and equal-angle resampling technique.The characteristic parameters in the signal order domain are input into the artificial neural network diagnosis model as fault sensitive features,and the fault types of rotor signals are classified.The practical application effect of the proposed method is verified by the measured signals,and the results are compared with those of traditional feature extraction method.The results show that the accuracy of the order analysis method is 99.8%and the standard deviation is less than 0.09%,which are better than the traditional feature extraction method.The rotor fault diagnosis method based on time-frequency ridge and order analysis has higher diagnostic accuracy,and has good feasibility and accuracy in the process of feature extraction of nonstationary signals.