Systems Engineering and Electronics ›› 2022, Vol. 44 ›› Issue (8): 2474-2482.doi: 10.12305/j.issn.1001-506X.2022.08.11

• Sensors and Signal Processing • Previous Articles     Next Articles

GMPHD based on measurement conversion sequential filtering for maneuvering target tracking

Zilin HOU, Ting CHENG*, Han PENG   

  1. School of Information and Communication Engineering, University of Electronic Science and Technology of China, Chengdu 611731, China
  • Received:2021-05-28 Online:2022-08-01 Published:2022-08-24
  • Contact: Ting CHENG

Abstract:

For multiple maneuvering targets tracking by Doppler radar in clutter, a multiple model Gaussian probability hypothesis density algorithm based on decorrelated unbiased converted measurement sequential filter is proposed. For the nonlinearity of the measurements, the position measurements are converted to unbiased measurements, and the Doppler measurement is converted to debiased pseudo measurement, and the tracking accuracy is improved by sequential filtering. For the maneuverability of the target, the idea of multiple model is introduced into Gaussian mixture probability hypothesis density (GMPHD), where the Gaussian components related to the model are predicted and updated. Simulation results demonstrate that the proposed algorithm can achieve effective maneuvering multi-target tracking in clutter. Compared with unscented Kalman multiple model GMPHD, the tracking accuracy is increased by 38.15%, and the algorithm efficiency is greatly improved. Compared with unscented Kalman best-fitting Gaussian approximation GMPHD, the efficiency is slightly increased, and the tracking accuracy is improved by 36.47%.

Key words: Gaussian mixture probability hypothesis density, multiple model, multiple target tracking, maneuvering target tracking, nonlinear measurement

CLC Number: 

[an error occurred while processing this directive]