Systems Engineering and Electronics ›› 2023, Vol. 45 ›› Issue (4): 982-990.doi: 10.12305/j.issn.1001-506X.2023.04.06

• Electronic Technology • Previous Articles    

PHD filter for small-angle crossing of multi-target trajectories

Zhengwei LIU, Ying CHEN, Yaobing LU   

  1. Beijing Institute of Radio Measurement, Beijing 100854, China
  • Received:2022-06-14 Online:2023-03-29 Published:2023-03-28
  • Contact: Ying CHEN

Abstract:

When multi-target trajectories cross at small angles, the spatially close targets and unresolved targets will lead to wrong estimation and track breakage in the probability hypothesis density (PHD) filter. To solve this problem, an improved PHD filter is proposed for small-angle crossing of multi-target trajectories. Firstly, a multi-target unresolved probability model is established, and the target detection probability is modified to adapt to the scene where there are unresolved measurements. Secondly, the weight of the updated Gaussian component is redistributed to ensure the one-to-one correspondence between spatial close targets and measurements. Simulation results show that the improved algorithm can effectively improve the target estimation accuracy and track continuity with small-angle crossing of multi-target trajectories, and it provide an effective solution to the difficult problem of multi-target trajectories crossing.

Key words: multi-target tracking, trajectory small-angle crossing, spatially close target, unresolved target

CLC Number: 

[an error occurred while processing this directive]