Feature Extraction of Impeller Fault Vibration Signal Based on Switched Kalman Filter
In order to improve the effective information extraction ability of fault signal time-domain waveform under filter,a switched Kalman filter algorithm was designed and applied to the feature extraction field of impeller fault vibration signal.The most likely state of monitoring data at all time points was predicted,the noise was removed and each impact component was effectively distinguished,and the signal to noise ratio was further strengthened.The simulation signal results show that the signal to noise ratio after filtering should be close to the noise ratio after adding noise,and the pulse discrimination effect is remarkable.The experimental verification results indicate that there is a significant noise in the measured signal,and the components of the signal judged at every moment are consistent with the reality.This research can be extended to other fields of mechanical transmission and has good market application value.
rolling bearingswitched Kalman filterfeature extractionnoise sign
国家科技期刊平台
NSTL
万方数据
