Analytical Algorithm for Attitude and Heading Estimation Aided by Maneuver Classification

Abstract

This paper presents a modified adaptive analytical algorithm for attitude and heading estimation. The analytical algorithm is based on the fusion of IMU, magnetometers and the velocity data from GPS. The kinematic Euler angles are first calculated based on the output of the rate gyros, then the calculated angle errors are compensated using the output of each of the accelerometers, magnetometers, and the velocity taken from a GPS receiver, without the need to model the systematic and random errors of the used sensors; Kalman filter is not used. The algorithm will be adaptive based on the maneuver classification, the filters’ parameters will be tuned depending on the maneuver intensity: No, Low, or High maneuver. The main contribution of this paper is to build an attitude and heading estimation algorithm (analytical algorithm) without using Kalman filter; this algorithm will be made adaptive based on the maneuver classification algorithm which was developed using logistic regression technique based on IMU output. Computer simulation with simulated and real flight data showed that the adaptive analytical algorithm has acceptable results compared to EKF.