An Improved Fading Kalman Filter and Its Application to GPS Kinematic Positioning

WANG Hu,WANG Jiexian,BAI Guixia,LI Haojun
DOI: https://doi.org/10.3969/j.issn.0253-374x.2011.01.024
2011-01-01
Abstract:Based on medium-long baseline GPS kinematic positioning model and the theory of classic fading Kalman filter,a modified fading factor vector is proposed.Different widths of sliding windows are used to obtain a series of parallel filters.The weight factor of covariance is decided by the residuals of every filter.The modified fading Kalman filter is more effective and sensitive than classic Kalman filter.Various experiments are made to test the performance of this method.The results show the precision of long-baseline GPS kinematic positioning is improved.
What problem does this paper attempt to address?