Navigation system of a class of underwater vehicle based on adaptive unscented Kalman fiter algorithm
来源期刊：中南大学学报(英文版)2014年第2期
论文作者：LIU Kai-zhou(刘开周) LI Jing(李静) GUO Wei(郭威) ZHU Pu-qiang(祝普强) WANG Xiao-hui(王晓辉)
文章页码：550 - 557
Key words：human occupied vehicle; navigation; extended Kalman filter; unscented Kalman filter; adaptive unscented Kalman filter
Abstract: Inherent flaws in the extended Kalman filter (EKF) algorithm were pointed out and unscented Kalman filter (UKF) was put forward as an alternative. Furthermore, a novel adaptive unscented Kalman filter (AUKF) based on innovation was developed. The three data-fusing approaches were analyzed and evaluated in a mathematically rigorous way. Field experiments conducted in lake further demonstrate that AUKF reduces the position error approximately by 65% compared with EKF and by 35% UKF and improves the robust performance.
LIU Kai-zhou(刘开周)1, LI Jing(李静)2, GUO Wei(郭威)1, ZHU Pu-qiang(祝普强)1, WANG Xiao-hui(王晓辉)1
(1. State Key Laboratory of Robotics (Shenyang Institute of Automation, Chinese Academy of Sciences),
Shenyang 110016, China;
2. Graduate University of Chinese Academy of Sciences, Beijing 100049, China)
Abstract:Inherent flaws in the extended Kalman filter (EKF) algorithm were pointed out and unscented Kalman filter (UKF) was put forward as an alternative. Furthermore, a novel adaptive unscented Kalman filter (AUKF) based on innovation was developed. The three data-fusing approaches were analyzed and evaluated in a mathematically rigorous way. Field experiments conducted in lake further demonstrate that AUKF reduces the position error approximately by 65% compared with EKF and by 35% UKF and improves the robust performance.
Key words:human occupied vehicle; navigation; extended Kalman filter; unscented Kalman filter; adaptive unscented Kalman filter
