Magnetic Field Aided Heading Estimation For Indoor Pedestrian Positioning

In the field of indoor pedestrian positioning, heading estimation plays an important role. Although gyroscopes are considered to be the primary sensors for attitude estimation, the errors associated with gyroscopes require periodic updates from other measurements. These measurements are magnetometers for the heading and accelerometers for roll and pitch estimates. The Quasi-Static magnetic Field (QSF) method has been proposed to estimate angular rate error in magnetically perturbed environments. However, this method may induce extra errors to the system due to the high false alarm probability in detecting quasi-static magnetic field. To address this issue, we propose a novel quasi-static magnetic detection approach for heading error estimation. The proposed quasi-static field detector detects the quasi-static field periods depending on the change rate of inclination angle, headings and the magnitude of the total magnetic field. In this manner, the detection accuracy and robustness are improved. Then the detected quasi-static magnetic field is utilized as the measurement fed into the ZUPT-aided EKF during the stance phase to estimate the heading errors. The experiment results indicate that the proposed method can provide superior performance in suppressing the heading errors with the comparison to original QSF method.