Research On Low-altitude Target Tracking Method

With the rapid development of electronic technology,navigation and control technology,more and more low-altitude flying targets are used widely.Due to the low flying of lowaltitude targets,there are a lot of false alarms in the measured data received by radar,and the low-altitude target motion has strong maneuverability,which brings great difficulties to the tracking of low-altitude targets.How to track low-altitude targets accurately and efficiently is an urgent problem to be solved.In this thesis,the performance of low-altitude target tracking algorithm is optimized from two aspects: improving the accuracy and realtime performance of target tracking.The main work is as follows:The data association algorithm based on FCM is mainly studied in this thesis.Firstly,aiming at the disadvantage of large amount of computation in JPDA algorithm,this thesis studies the JPDA algorithm based on FCM(FJPDA).The association probability is calculated by calculating membership degree through FCM clustering,the separation of confirmation matrix is avoided,and the computation is reduced to a great extent in this thesis.At the same time,the algorithm of NN based on FCM(FNN)is studied.The algorithm performs nearest neighbor association by FCM clustering,which is simple in computation and can better correlate targets with high false alarm in measurement data.The simulation results show that the FJPDA and FNN algorithms can reduce the computational complexity to a great extent,but their correlation accuracy decreases slightly.In order to improve the correlation accuracy,the idea of distance quadratic weighting is presented to improve the FJPDA algorithm(IFJPDA),and the distance weights of common measurements are attenuated,and the correlation probability is adjusted by distance weights in this thesis.The simulation results show that the IFJPDA algorithm can reduce the computational complexity and ensure high correlation accuracy.The adaptive CS algorithm is mainly studied in this thesis.The CS algorithm has the disadvantage of poor tracking effect when tracking weak maneuvering or non-maneuvering targets or when the acceleration of the target suddenly changes.In this thesis,an adaptive CS algorithm based on the amendment of the acceleration membership function is studied.The algorithm modifies the acceleration extremum through the membership function,adjusts the acceleration extremum adaptively according to the actual acceleration of the target,and introduces the amendment factor to improve the tracking delay when the acceleration changes suddenly.Then an adaptive CS algorithm based on membership function modification of innovation is presented.The acceleration extremum is modified by innovation and the tracking accuracy is improved by modifying the predicted value in this thesis.Finally,considering the influence of maneuvering frequency on CS algorithm,the IMM algorithm based on adaptive CS model is also analysed in this thesis.Because the two adaptive CS models with different maneuvering frequencies can more fully represent the actual motion characteristics of the target,thus effectively improving the tracking accuracy.The simulation results show that the adaptive CS algorithm modified by membership function and the IMM algorithm based on adaptive CS can effectively reduce tracking error when tracking weak maneuvering or non-maneuvering targets and when the acceleration of the target suddenly changes.