I.A. Nagin

GNSS/IMU integration is the typical solution for highly dynamic objects navigation. However, the use of a Gaussian approximation of the theory of optimal filtering (leading to EKF) in an highly nonlinear problems does not guarantee a good performance. Therefore, in the last decade intensively developed alternative approaches to the problem of filtration of system process and/or observations described by highly nonlinear models. One of these approaches is named as the Unscented Kalman Filter (UKF). In this paper we investigate the efficiency of the UKF approach in the integration problem by comparing it with the classical EKF algorithm.