Heading angle estimation using rotating magnetometer for mobile robots under environmental magnetic disturbances

The heading angle plays a vital role in the localization and mapping of mobile robots. It is generally obtained by fusing measurements from gyroscope and magnetometer. However, ferromagnetic objects in real-world environments will disturb the magnetic field and will, therefore, cause significant errors in the estimated heading angles. This work proposes a novel method that employs a rotating magnetometer to detect ambient spatial magnetic disturbances and corrects the heading angle. The algorithm is based on the extended Kalman filter (EKF). Firstly, a criterion named spatial disturbance index is defined to characterize the disturbance quantitatively. And then the magnetometer measurement error covariance of the EKF is tuned adaptively according to the proposed criterion, so that a relatively reliable heading angle can be obtained even under strong spatial dynamic magnetic disturbances. In addition, the estimated heading angle can quickly restore to the correct value when the spatial disturbances disappear. The proposed algorithm has the benefit of adjusting the fusing degree of gyroscope and magnetometer adaptively to reject spatial disturbances and avoid the adverse impact of inherent gyroscope drift. The algorithm is evaluated under static and dynamic conditions in real-world indoor/outdoor environments. The results show that our algorithm outperforms the conventional EKF with fixed measurement error covariance and also the algorithm using only gyroscope.