Research On Radar Maneuvering Target Tracking Algorithm Based On Interacting Multiple Model

With the rapid development of science and technology,the maneuvering performance of modern target is more and more strong,and the tracking performance of radar is also put forward higher requirements.So the radar maneuvering target tracking technology has more extensive development prospects.In the field of target tracking,the tracking of maneuvering target is always a difficult problem.In order to track the maneuvering target stably and accurately,this thesis studies the maneuvering target tracking problem based on interacting multiple model(IMM)algorithm.Firstly,the basic principle of maneuvering target tracking was analyzed,and several common target motion models were discussed.Then the extended Kalman filter(EKF)algorithm and unscented Kalman filter(UKF)algorithm were analyzed,and the principles,advantages and disadvantages of these two nonlinear filtering algorithms were studied respectively.The principle and characteristics of the IMM algorithm were mainly studied,and combined IMM algorithm with EKF algorithm and UKF algorithm respectively.The tracking performance of IMM-EKF algorithm and IMM-UKF algorithm was compared by simulation.The simulation results show that the tracking performance of the IMM-UKF algorithm is better.Secondly,the factors that affect the performance of IMM algorithm were studied,and an adaptive parallel IMM algorithm based on UKF(APIMM-UKF)was proposed.The influence of transition probability matrix(TPM)on the tracking performance of IMM algorithm was studied.In order to further improve the tracking performance of the IMM-UKF algorithm,the TPM in this algorithm was adaptively processed to obtain the IMM-UKF algorithm based on TPM adaptation(AIMM-UKF),and its performance was simulated and compared.Then,in response to the problem of increased tracking peak error in the AIMM-UKF algorithm,the AIMM-UKF algorithm and the IMM-UKF algorithm were run in parallel,and the two algorithms were integrated into a unified whole through the transition probability correction function to completed the design of APIMM-UKF.The tracking performance of APIMM-UKF was simulated and verified.The results show that,compared with the AIMM-UKF algorithm,the APIMM-UKF algorithm improves the model switching speed and reduces the tracking peak error caused by the model switching.Finally,in order to further reduce the tracking peak error and improve the tracking accuracy of maneuvering targets,an improved APIMM-UKF algorithm was proposed.This chapter first studied the IMM interactive mixing method based on the unequal dimension state,analyzed the influence of different extended dimension interaction methods on the performance of the IMM algorithm.The performance of IMM algorithm based on unequal dimension state interaction was simulated and analyzed.The simulation proved that the use of unequal dimension state expansion interaction method can effectively reduce the tracking peak error caused by model switching.Then an improved APIMM-UKF algorithm was proposed on this basis,which used the unequal dimension state expansion interaction method to improve the interaction process in the APIMM-UKF algorithm.The tracking performance of the improved APIMM-UKF algorithm was simulated and verified.The results show that the improved APIMM-UKF algorithm improves the model switching response speed,reduces the peak error,and meets the needs of maneuvering target tracking.