Robust Sensor Fusion for MEMS GPS/IMU in Diverse Flight Environments

Jose A. Rios, Darren Liccardo

Abstract:	A low cost solid state GPS/IMU navigation unit has been developed that incorporates measurements from GPS, MEMS gyros and accelerometers, and fluxgate magnetometers to provide a navigation solution at a high output rate. The Crossbow Technology, Inc. AHRS500 and NAV420 family of inertial sensor products provide stand alone solutions for vertical gyro applications, attitude and heading reference system (AHRS) applications, and navigation GPS/IMU applications. Firmware inside the onboard processors produces calibrated angular rate measurements, calibrated acceleration measurements, calibrated magnetometer measurements, and the estimated navigation state which includes body attitude (roll, pitch, heading), local level horizontal navigation frame position (latitude, longitude, and altitude) and velocity. The algorithm used to estimate the navigation state is an Extended Kalman Filter (EKF) trajectory correction approach in which the inertial accelerometers and gyros propagate the state trajectory made up of velocity, and body attitude, and the supporting sensors (GPS and magnetometers) provide velocity and earth magnetic field measurements which the filter uses to calculate corrections to the trajectory state, and estimate inertial sensor errors and ferrous material effects. Applications in Unmanned Aerial Vehicles (UAV) and General Aviation (GA) require robustness in ways usually not required by specialized navigation systems; (1) the navigation solution must degrade gracefully under a variety of sensor failures, and (2) the system must provide an accurate navigation solution under a variety of flight profiles, some highly dynamic and others mostly static. Each of these robustness requirements is addressed with a variety of heuristic enhancements to manipulate the EKF covariance parameters, filter design, and measurement models. We present performance results under a variety of flight test profiles. Under static conditions, the maximum attitude and heading errors are typically less than 0.5 degrees, and under dynamic flight tests flown against a traditional high accuracy INS system (Litton LN-100G), the attitude and heading RMSE are shown to be less than 0.4 degrees and 1.8 degrees, respectively. The position and velocity estimates directly follow the GPS accuracy level when the GPS is providing a low GDOP solution. When the GPS accuracy level drops due to satellite occlusion, the combined solution maintains the accuracy level when compared to the INS and smoothes over GPS estimate error nonlinearities. Under prolonged GPS failure the system degrades to AHRS functionality using the EKF estimated magnetometer calibration for true heading information. Under poor magnetometer health, the system further degrades to a stable attitude estimator. The added robustness achieved by heuristic sensor fusion techniques and the traditional EKF provide a flexible navigator that can be deployed in a variety of UAV and GA applications. The miniaturization achieved by the NAV420 product achieves this functionality in a fully environmentally sealed unit approximately three inches on a side.
Published in:	Proceedings of the 17th International Technical Meeting of the Satellite Division of The Institute of Navigation (ION GNSS 2004) September 21 - 24, 2004 Long Beach Convention Center Long Beach, CA
Pages:	833 - 844
Cite this article:	Rios, Jose A., Liccardo, Darren, "Robust Sensor Fusion for MEMS GPS/IMU in Diverse Flight Environments," Proceedings of the 17th International Technical Meeting of the Satellite Division of The Institute of Navigation (ION GNSS 2004), Long Beach, CA, September 2004, pp. 833-844.
Full Paper:	ION Members/Non-Members: 1 Download Credit Sign In