Comparision of Kalman Filter and Unscented Kalman Filter to Estimate a Non-Linear Systems Position

Abstract

Estimating a real time location of an Autonomous Underwater Vehicle (AUV)'s is always of great interest. This paper discusses the use of Kalman filters to control AUV's. The Kalman filter (KF) is an ideal linear estimator when white Gaussian noise can be used to model the system noise and the measuring noise. But Unscented Kalman Filter has been proven to be superior alternative to Kalman Filter. So in order to cope up with the non-linearity of the system Unscented Kalman Filter has been adopted.There are two facets to the protocol:1)first, tracking of AUV using a kalman filter. 2)second ,tracking using UKF.The paper offers a contrast for a nonliner model between Kalman Filter and Unscented Kalman Filter. The results of the simulations demonstrate the efficacy of the proposed tracking scheme, error the estimated displacement and the displacement measured.