Research On The Application Of Tracking And Filtering Methods In Interception System

Precision-guided weapons play an increasingly important role in modern military warfare,gradually becoming an important measure of a country’s military strength.The key part of precision guidance technology is the high-performance tracking and filtering technology.Based on the missile interception system,this paper focuses on the tracking and filtering methods for maneuvering targets.Firstly,based on the three most commonly used maneuvering target models,CA model,Singer model and “current” statistical model,the Kalman tracking filters using ground radar ranging information are designed respectively and the corresponding simulation experiments are set under the same condition.The performance differences of the three types of models for maneuvering target tracking are analyzed.The simulation results show that the "current" statistical model has a good advantage in tracking maneuvering target,but it shows a certain disadvantage in tracking non-maneuvering motion.Secondly,the Gaussian Mix Probability Hypothesis Density(GM-PHD)multi-target tracking method based on stochastic finite set theory is expounded.And the detailed algorithm steps are given.Based on seeker actively ranging.information,the GM-PHD tracker is designed under linear Gaussian model.And the multi-to-multi target tracking simulation experiments under ideal condition and clutter environment have been carried out respectively,which have verified the multi-target tracking performance of GM-PHD in multi-to-multi interception system.Finally,according to the principle of extended Kalman,GM-PHD is extended to nonlinear systems.Based on the extended GM-PHD filtering algorithm,a multi-target tracker with measurements of LOS angles is designed and the multi-target passive tracking is performed in multi-to-multi interception system.Simulation experiments have been carried out to verify the performance of the extended GM-PHD multi-target tracking method.The simulation results under both linear and nonlinear systems show that the GM-PHD filtering method can not only achieve high-precision tracking of multi-maneuvering target states,but also effectively eliminate clutter interference and achieve accurately estimating of the number of targets in the field of view.