Abstract:

Aiming at the weakly observability and highly nonlinearity of a single observer of infrared search and track (IRST) systems, a multi-observer fusion tracking algorithm based on modified iterated extended Kalman filter (MIEKF) is proposed. The IEKF is modified by providing a new measurement update with Gauss-Newton iteration algorithm, then an iterative termination condition is deduced based on a maximum likelihood criterion, thus the linearity error is reduced. Finally the MIEKF combining with the central fusion tracking algorithm is applied to multi-observer target tracking of IRST. Simulation results show that the proposed algorithm is better than EKF and UKF for a three-observer target tracking system.