Multiple adaptive notch filters based a time-varying frequency tracking method for Coriolis mass flowmeter

Abstract

A method based on multiple adaptive notch filters for tracking the frequency variations of the sensor output signal of Coriolis mass flowmeter is presented. For the simulated signal whose frequency, amplitude and phase are time-varying based on the random walk model, a lattice-based adaptive IIR notch filter (LANF) with fast convergence rate is applied to detect the signal frequency and track the time-varying frequencies initially. As the LANF algorithm converges, a simplified lattice-based adaptive IIR notch filter (SLANF) sets its initial notch frequency which equals to the value detected by LANF, and then starts to work. After the SLANF algorithm converges, the time-varying frequencies are tracked by SLANF instead of LANF. Computer simulations show that the presented method can attain a favorable compromise between convergence rate and tracking accuracy, and its calculation complexity is lower than the LANF method merely.