Adaptive Precise Attitude Estimation Using Unscented Kalman Filter in High Dynamics Environments

Abstract

In this paper, a novel adaptive scaled unscented Kalman filter (ASUKF) algorithm is developed using low-cost micro-electro-mechanical system (MEMS) triaxial gyroscope and accelerometer. The body non-gravitational accelerations are estimated and used to compensate the accelerometers measurements. The estimated external acceleration is used to adapt the acceleration measurement noise covariance matrix to achieve robustness during harsh environments. The proposed ASUKF uses the adapted covariance matrix and the compensated accelerometer measurements to precisely estimate the body attitude angles. The achieved accuracies for the proposed model are discussed and compared with other state-of-the-art algorithms through a laboratory and field tests. The results show that the proposed algorithm achieves an outstanding level accuracy in high dynamics environments in comparison to other attitude estimators.