Kalman filtering to real-time trace water level measurements using ultrasonic sensor

Abstract

Water level monitoring systems are needed in various related fields. However, to accurately determine water levels typically difficult due to the effect like sloshing in dynamic environments. In this paper, we proposed Kalman filter based approach to reduce uncertainty in the water level measurements. Experiments have been performed by filling the tank with water and installing an ultrasonic sensor module at the top center of the tank. The water tank is vibrated with an artificial vibrator, then measuring the water level is done by the ultrasonic sensor. To reduce measurement errors we introduced the Kalman filtering. This technique is developed for improving the precision of water level measurements in the presence of various noise sources. The Kalman filter performance is demonstrated to be adaptive to real-time noise. Experiment results indicate reducing the errors of measurement significantly up to about 60% for a dynamic water condition by adjusting appropriately the Kalman filter tuning parameters.