Dynamic offset correction of electromagnetic flow meters

This contribution addresses the electrochemical voltage drift of electromagnetic flow meters. We will extend the conventional discrete time alternating field method to continuous operation. A Kalman filter, based on a state space model of the sensor, was developed to separate the flow-proportional signal from the noisy voltage drift of the electrodes. The new algorithm allows to process flow information rarely used today, i.e., the electrode voltage during times of stable magnetic field. Thereby a lower energy consumption and an increased accuracy can be achieved. Compared to conventional signal processing, the performance of the new method is far more independent of the transient behavior of the actual flow. Measurements show a comparable accuracy, but with a significantly reduced energy demand. Off-line signal processing let expect a theoretical power reduction factor around 150.