Open AccessTarget tracking algorithm based on adaptive strong tracking particle filter
Author(s) -
Jiaqiang Li,
Ronghua Zhao,
Jinli Chen,
Chunyan Zhao,
Yanping Zhu
Publication year - 2016
Publication title -
iet science, measurement and technology
Language(s) - English
Resource type - Journals
SCImago Journal Rank - 0.418
H-Index - 49
eISSN - 1751-8830
pISSN - 1751-8822
DOI - 10.1049/iet-smt.2016.0044
Subject(s) - tracking (education) , particle filter , control theory (sociology) , auxiliary particle filter , algorithm , stability (learning theory) , residual , computer science , moment (physics) , filter (signal processing) , adaptive filter , tracking system , kernel adaptive filter , mathematics , kalman filter , filter design , ensemble kalman filter , artificial intelligence , extended kalman filter , computer vision , physics , psychology , pedagogy , control (management) , classical mechanics , machine learning
The primary problem of tracking filtering algorithms is the tracking stability and effectiveness of target states. Based on the particle filter, an adaptive strong tracking particle filter algorithm is proposed in this study. According to the residual between actual measurement values and predicted measurement values of every moment, adjustment of the forgetting factor and the weakening factor is adaptively conducted. Then, by calculating the fading factor, transfer covariance matrix and filter gain of the system are obtained to estimate the particles state value. Updating the importance density function can alleviate the degradation phenomenon of particle filter, and it contributes to effective estimation for the optimal state value of a target. The simulation results demonstrate that the proposed algorithm provides a better tracking precision. In addition, when the target states make mutations, the proposed algorithm can track the mutation states of moving targets effectively and improve the stability of the system.