Research On Multi-target Tracking Algorithm Based On Multi-source Information Fusion And Random Finite Set

With the development of modern radar,photoelectric,sonar and other sensor technologies,multi-target tracking(MTT)technology has become one of the research hotspots in the field of information fusion,and is widely used in military and civilian fields such as air early warning,traffic command and so on.In the modern battlefield environment,the uncertainty factors such as high-density background clutter,the number of targets changing with time,and the complex maneuvering of multiple targets will lead to the increase of the calculation amount of the traditional multi-target tracking algorithm,and the decrease of the timeliness and tracking accuracy of the algorithm.Random Finite Set(RFS)algorithms do not need complex data association operations,and can be effectively combined with multi-source information fusion algorithms,providing a new way to solve the problem of multi-target tracking.this dissertation based on multi-source information fusion technology and stochastic finite set algorithm,the multi-target tracking problem is studied in depth.The main research contents are as follows:1.Aiming at the problem of tracking accuracy degradation caused by high background clutter and complex maneuvering of target in linear Gaussian scene,a multi-model Gaussian mixture potential probability hypothesis density algorithm based on Doppler information joint filtering and sequential filtering is proposed.Based on the Cardinality Probability Hypothesis Density(CPHD)filter,the algorithm first designs a dual-correlation gate based on position-Doppler information,and uses the position information and Doppler information to jointly filter the measurements,which improves the accuracy of filtering;Secondly,the multi-model algorithm is introduced to match the multiple maneuvering modes of the target;Finally,based on the correlation between the target Doppler information and the position information,the Doppler information is used for sequential filtering on the basis of the position update results to further improve the tracking accuracy of multiple targets.Simulation results show the effectiveness of the proposed algorithm.2.Aiming at the Doppler blind zone problem of radar sensors in nonlinear and low clutter density scenarios,a multi target tracking algorithm based on distributed information fusion of multiple sensors is proposed using the complementary information characteristics of multiple sensors.The algorithm is based on a fifth order CKF potential balanced multi target multi Bernoulli(CBMe MBer)filter,utilizing the performance advantage of passive sensors that are not affected by Doppler blind zone.A target tracking method within the Doppler blind zone is designed,and the information complementarity between multiple sensors is utilized to construct the 5th CKF CBMe MBer algorithm based on distributed multi-source information fusion outside the Doppler blind zone.Simulation experiments show that the proposed multi-sensor information fusion algorithm can achieve target tracking within the Doppler blind zone and improve the accuracy of multi target tracking outside the Doppler blind zone.