Research And Implementation Of Target Tracking Algorithm On Passive Radar Reconnaissance Platform

Today,with the rapid development of the information age,the modernization of the military is also constantly improving.In the face of an increasingly complex electromagnetic environment,the drawbacks of traditional radar reconnaissance platforms are undoubtedly revealed.Generally speaking,the traditional radar reconnaissance platform has a low degree of complexity and only pays attention to the functions of the respective parts of each module.The radar reconnaissance platform has a low integrity,and in many cases cannot meet the demand for a unified radar reconnaissance platform in modern warfare.Second,the radar target tracking process plays an important role in the radar reconnaissance platform.After receiving the processed radar signal parameters,whether it can predict the target trajectory,track the target in real time,eliminate irrelevant interference,etc.,are all radar target tracking Content of the study.Among them,the tracking of multiple targets is even more important and difficult in the field of radar targets.Aiming at these two problems,this paper designs and develops a complete radar reconnaissance platform based on target tracking,and at the same time conducts in-depth research on multi-target tracking.This paper studies the classical Probability Hypothesis Density algorithm and analyzes its advantages and disadvantages.On this basis,Gaussian Mixture Probability Hypothesis Density filtering is realized through simulation,and an improved GM-PHD algorithm is proposed for the defect of target number estimation revealed by the GM-PHD algorithm.The results show that the improved algorithm has a significant improvement in the accuracy and accuracy of target tracking under the premise of guaranteeing time.The content and results of this paper are as follows:First of all,this paper conducts simulation analysis on the principle of single target tracking algorithm in different environments.The Kalman Filter series algorithms are studied separately,and they are combined with the Particle Filter algorithm.On the basis of the Extended Kalman Filter and the Unscented Kalman Filter,the Extended Particle Filter and the Unscented Particle Filter are realized.At the same time,the principle of Random Finite Set theory is introduced.The single-target Kalman filter algorithm is simulated in different application scenarios.Through the simulation results,the advantages and disadvantages of the above-mentioned target tracking algorithm in different environments and the adaptation conditions are discussed.Secondly,the principle of Probability Hypothesis Density algorithm is studied,which uses the first moment of the posterior probability distribution of target tracking to replace the calculation of probability density.After that,Gaussian Mixture Probability Hypothesis Density filtering is introduced.However,GM-PHD filtering also has its own shortcomings.When the tracking conditions are more disturbed,the GM-PHD filtering may not accurately estimate the number of targets,leading to errors.In response to this problem,this paper proposes an improved GM-PHD algorithm,which avoids the error estimation problem in the GM-PHD algorithm by modifying the traditional GM-PHD.Finally,this article develops a complete set of radar reconnaissance software experimental platform,which includes radar sorting,radar recognition,radar direction finding,radar positioning,radar tracking and other modules in the radar reconnaissance system,and has a good human-computer interaction interface.It can quickly call the algorithm and display the simulation results.