Research On Radar Weak Target Track-Before-Detect Algorithm Based On Dynamic Programming

With the diversification of targets and the complexity of the actual environment,the detection capabilities of modern radars are facing very serious challenges,and the detection and tracking of weak targets is one of them.Tracking before detect(Track-Before-Detect,TBD)is a method for effectively detecting weak targets in a low signal-to-noise ratio environment,and track-before-detect algorithm based on dynamic programming(DP-TBD)is the main implementation of TBD technology.The DP-TBD algorithm can transform a more complex multi-level decision problem into multiple single-level decision processes,and then optimize each single-level decision to obtain a global optimal solution.The algorithm has clear principles and excellent performance,so it is a research hotspot in recent years.This article will focus on the DP-TBD algorithm for radar weak targets.The main contents are summarized as follows:(1)Analyzing the basic theory of dynamic programming and apply it to the DP-TBD algorithm,then build the main system model of the traditional DP-TBD algorithm,and introduce the implementation process of the DP-TBD algorithm under the model framework The merit functions and state transition rules that affect the performance of the algorithm are specified.(2)Focusing on the study of several single-target DP-TBD algorithms and their improved methods,and constructed a measurement index to evaluate the performance of the algorithm.Aiming at the problems of the energy diffusion phenomenon and the poor tracking effect of maneuvering targets in the traditional DP-TBD algorithm,the DP-TBD algorithm based on directional weighting,the DP-TBD algorithm with adaptive state transition set,and the DP-TBD algorithm based on posterior probability density accumulation are proposed.Finally,the experimental simulations of these DP-TBD algorithms were carried out,and the results confirmed the effectiveness of these improved algorithms.(3)Analyzing the main technical difficulties of the multi-target detection and tracking algorithm at first,and the system model of the multi-target DP-TBD algorithm(MT-DP-TBD)is established on the basis of the single-target algorithm.Then according to the distribution of the range of motion of each target in the observation scene,the DP-TBD algorithm when multiple targets are far away from each other is studied at first,then the DP-TBD algorithm under the arbitrary distribution of multiple targets is discussed,and finally constructed evaluate the indicators and design simulation experiments to verify the effectiveness of these two types of DP-TBD algorithms.