We present a filtering algorithm for processingout-of-sequencemeasurements (OOSMs) in multi-sensor target tracking using the particle filter (PF). measurements can arrive out-of-sequence at a fusion center due to va...
详细信息
ISBN:
(纸本)078037231X
We present a filtering algorithm for processingout-of-sequencemeasurements (OOSMs) in multi-sensor target tracking using the particle filter (PF). measurements can arrive out-of-sequence at a fusion center due to varying data pre-processing times at the sensor platforms and communication delays in a multi-sensor target tracking system. Standard filtering algorithms such as the Kalman filter (KF), extended KF (EKF), interacting multiple model (IMM) filter, and PF can not be directly applied to process the OOSM. processing of the OOSM requires new algorithms within the framework of KF, EKF, IMM, or PF. The ground moving target indicator (GMTI) radar sensors play an important role in military and commercial domain such as surveillance of ground moving targets and air traffic control. The GMTI measurement model is a nonlinear function of the target state. Therefore, extension of the existing linear OOSM filtering algorithms to the GMTI filtering problem with OOSMs would require extended Kalman filter (EKF) type linearization and hence would be sub-optimal. As a superior alternative to the EKF, the PF has been successfully applied to a number of nonlinear filtering problems where the EKF or IMM using EKF is sub-optimal. Our OOSM PF algorithm does not require a large amount of storage and preliminary results based on numerical simulations show that the algorithm yields comparable results as previous algorithms.
暂无评论