Comparison and error analysis of integral-free Kalman tracking filter algorithms

Author

Hongyan Wang ; Daobin Yu ; Jiawei Jiang

Author_Institution

Dept. of Inf. Equip., Acad. of Equip., Beijing, China

fYear

2014

fDate

14-16 Oct. 2014

Firstpage

783

Lastpage

787

Abstract

The integral-free Kalman filters which are widely used in target tracking are studied. The algorithms of Unscented Kalman filter (UKF), Cubature Kalman filter (CKF) and Square-root cubature Kalman filter (SCKF) are compared in details. A modified algorithm (MSCKF) is proposed to optimize the performance. When considering different original ranges and radial speeds, simulation of linear motion targets in Gauss noise is made and their tracking errors are analyzed. The result shows that different tracking filter algorithm has respective features in short time and long range signal processing. The MSCKF has better tracking performance in short time tracking application. It offers the guideline for application.

Keywords

Gaussian noise; Kalman filters; error analysis; filtering theory; least mean squares methods; nonlinear filters; recursive estimation; target tracking; Gauss noise; SCKF; UKF; cubature Kalman filter; integral-free Kalman tracking filter algorithms; linear motion targets simulation; long range signal processing; minimum mean square error; modified algorithm; recursive MMSE estimator; short time signal processing; square-root cubature Kalman filter; target tracking; tracking error analysis; tracking filter algorithm; unscented Kalman filter; Filtering algorithms; Kalman filters; Noise; Noise measurement; Radar tracking; Target tracking; MSCKF; SCKF; integral-free Kalman tracking filter;

fLanguage

English

Publisher

ieee

Conference_Titel

Image and Signal Processing (CISP), 2014 7th International Congress on

Conference_Location

Dalian

Type

conf

DOI

10.1109/CISP.2014.7003883

Filename

7003883