Signal decoupling and analysis from inner flush-mounted electrostatic sensor for detecting pneumatic conveying particles

Abstract Comparing the frequency spectrums of the measured signals from both outer surface-mounted electrostatic sensor and inner flush-mounted electrostatic sensor, it is found that the transferred charge signal which used to be ignored appears primarily in the mean value of the measured signal. Furthermore the signal frequency spectrum of inner flush-mounted electrostatic sensor contains a critical frequency for decoupling induced charge signal and transferred charge signal. Then harmonic wavelet transform is proposed to decompose the measured signal. As a result, the induced charge signals from the upstream and downstream electrodes have higher similarity than the directly measured signals. Thus the mean cross-correlation coefficient increases from 0.61 to 0.78, which contributes to the cross-correlation velocity. In addition, the RMS value of the transferred charge signals increases linearly with the increasing of both the particle mass flow rate and the superficial air velocity, which is an effective prediction method for particle mass flow rate.