Open AccessGaussian sum pseudolinear Kalman filter for bearings‐only tracking
Author(s) -
Jiang Haonan,
Cai Yuanli
Publication year - 2020
Publication title -
iet control theory and applications
Language(s) - English
Resource type - Journals
SCImago Journal Rank - 1.059
H-Index - 108
eISSN - 1751-8652
pISSN - 1751-8644
DOI - 10.1049/iet-cta.2019.0597
Subject(s) - observability , kalman filter , control theory (sociology) , extended kalman filter , invariant extended kalman filter , mathematics , tracking (education) , gaussian , ensemble kalman filter , metric (unit) , fast kalman filter , filter (signal processing) , alpha beta filter , algorithm , computer science , engineering , artificial intelligence , computer vision , physics , psychology , pedagogy , control (management) , quantum mechanics , moving horizon estimation , statistics , operations management
The efficacy of a bearings‐only tracking algorithm, to a great extent, depends on the target‐sensor geometry and motion. Although the pseudolinear Kalman filter and its variants have demonstrated comparable performance with the elite non‐linear filters, they still suffer from bias problems and the tracking performance is inevitably affected by the relative geometry and motion relationships. In this study, an observability metric based on classical control theory is first presented to characterise the relative relationships between the target and the sensor. Then an efficient bearings‐only tracking algorithm called Gaussian sum pseudolinear Kalman filter is developed. It is based on the bias‐compensated pseudolinear Kalman filter and is built within a Gaussian sum framework. In the novel algorithm, a splitting and merging procedure will be triggered when a low degree of observability is detected. Simulation results show the significant performance improvement of the proposed algorithm.