Oil metal abrasive swarf detection system based on Kalman filtering

LIU Wei,WEN Peng,ZHOU Shichao,SUN Jiangang,LI Kai

DOI: https://doi.org/10.3969/j.issn.1673-1131.2023.03.012

2023-01-01

Abstract:Faced with the detection needs of metal abrasive chips in oil, aiming at the problem that the detection accuracy of traditional oil metal abrasive swarf detection systems is easily affected by noise, an oil metal abrasive swarf detection system based on Kalman filtering is designed. Firstly, the magnetic excitation sensor is designed according to the principle of electromagnetic induction and the voltage disturbance model is established. Secondly, the Kalman filter is designed according to the characteristics of the metal abrasive chip voltage signal. Thirdly, the voltage disturbance model is verified by data fitting and the process noise covariance is calculated. Then the measured noise covariance is obtained by wavelet analysis. Finally, according to the obtained parameters, the metal grinding chip voltage disturbance data is Kalman filtered to obtain the final result. The test results show that after Kalman filtering, the voltage disturbance signal error of300um metal grinding chips is reduced by 2%, the voltage disturbance signal error of 400um metal grinding chips is reduced by 2.5%, and the voltage disturbance signal error of 500um metal grinding chips is reduced by 3%, which proves that Kalman filter is feasible for this system.