Research On Tracking Before Detection Algorithms For Latticeless Dynamic Programming

Radar echo signals reflected by low observable targets such as dim and small targets are greatly weakened,and the background environment of targets is becoming more and more complex.Target maneuverability also increases the difficulty of radar detection and easily causes target loss.Therefore,the diversity of targets and the complexity of the environment put forward higher requirements for the detection ability and track processing of modern radar.With the rapid development of radar technology,the detection and tracking of weak targets has also received extensive attention.Compared with the traditional detection before tracking method,the pre-detection tracking technology is more efficient and has become a hot research topic.TBD technology can greatly reduce false tracks and improve the detection and tracking performance of weak targets.The main work of this paper is as follows:(1)Comparisons and analyses are made between detection before tracking and pre-detection tracking algorithms.Discussing the similarities and differences of the four methods based on three-dimensional matched filter,projection transform,multilevel hypothesis testing,higher-order correlation,particle filter and dynamic programming.(2)The pre-detection tracking algorithm based on dynamic programming is emphatically studied and analyzed.The target motion model and measurement model based on dynamic programming algorithm,the DP-TBD state transition model and the selection of value function are built.The flow chart of DP-TBD algorithm is analyzed,and the evaluation index and performance index are deduced.The accumulation of value functions under different frame numbers and different SNRs is simulated and compared.(3)Discusses the limitations of traditional pre-detection tracking algorithm based on dynamic programming.In order to improve these shortcomings,an improved algorithm of lattice-free DP-TBD is proposed.The basic idea and algorithm flow are described in detail,the applicable model is analyzed,and the motion and measurement model is given.The proposed algorithm is simulated and analyzed to test whether the new algorithm has better performance.