Quadratic EKF algorithm enhancements for low cost tightly-coupled AHRS/GPS

摘要

Attitude and heading reference system (AHRS) is a three dimensional measurement package including micro-inertial measurement units (MIMUs) and magnetic sensors, which can accomplish the real-time attitude control of moving bodies. When GPS is added to an AHRS, the combined system becomes a micro inertial sensors based INS/GPS. The whole system, in contrast, can output the position, velocity and attitude information in chosen coordinate frames. The tightly-coupled measurement mode that fuses pseudo_range-pseudo_range rate-heading measurements is proposed and intensively illustrated. Due to the large drifts of MIMUs, the quadratic extended Kalman filter (QEKF) algorithm is introduced substituting for conventional EKF, compensating the quadratic truncation error extracted from parameter linearization according to Taylor expanding theory. The simulation under the same condition indicates that the accuracy of the navigation parameters is highly enhanced, better than the ones obtained from the standard EKF scheme.