Huang, H.; Chen, X.; Zhou, Z., and Lv, C., 2015. Attitude determination for underwater gliders using unscented Kalman filter based on smooth variable algorithm.
For improving the attitude accuracy using the lower cost and lower power underwater navigation system, this paper proposes the new underwater navigation system which is composed of the inertial sensors aided with the magnetometer, and the unscented Kalman filter based on smooth variable algorithm (UKF-SV) is proposed to improve the attitude accuracy for the underwater glider. The UKF-SV makes use of the advantage of UKF (unscented Kalman filter) in the nonlinear model to estimate attitudes and then smoothes the variables estimated by UKF. Through this process, the attitude accuracy can be greatly improved. The convergence of UKF-SV is proven in theory. The static and dynamic experiments are done to assess the performance of UKF-SV and compare with the traditional UKF. The experiment results show that the performance of UKF-SV is better than the traditional UKF.