A comparison of DVL/INS fusion by UKF and EKF to localize an autonomous underwater vehicle

In this paper, the position of an autonomous underwater vehicle (AUV) has been estimated by fusion of the data of two sensors: Doppler velocity log (DVL) and inertial navigation system (INS). Two different filters have been used in order to estimate the position of AUV, namely, Extended Kalman Filter (EKF) and Unscented Kalman Filter (UKF). The approach of EKF is based on linearization of system state space equation around the instantaneous values of the state variables. The UKF is based on the nonlinear transformation of selected points of a Gaussian distribution of the state variables. These two filters are implemented, using nonlinear kinetic model of a sample AUV, and the results of the position estimation of the two filters are compared. The results show that despite the linearization approximations, the EKF results are closer to the real path of the vehicle than the UKF estimates.