Research On Motion Target Tracking Algorithm Based On GNSS

Nowadays,with the continuous development of information science and technology and the informatization,intelligence and integration of modern warfare,the motion target tracking application based on the Global Navigation Satellite System(GNSS)has been widely used.It involves the military field and the civilian life field.In the military field,such as missile precision guidance,air missile defense and attack,and ship cruise,etc.;in the civil field,thereare traffic control,aviation navigation and positioning,and anti-collision systems for port ships.With the emergence of various new military sports targets and the increasing demand for accuracy,real-time and reliability of tracking information in life,motion target tracking based on GNSS system has become one of the current research hotspots.This thesis mainly carries out the research on the motion target tracking algorithm based on the GNSS system.The pseudo-distance positioning equation and the Doppler velocity equation constitute the GNSS navigation model,and the equation composed of the two is also the measurement equation in the Kalman filter algorithm in the motion target tracking process,so for the Kalman filter algorithm Research is inevitable.In the current research process,when the performance error of the model or the subsequent measurement abnormality occurs,the traditional cubature Kalman filter(CKF)algorithm will still have problems such as significantly reduced filtering performance or divergence of filtering effect,and In the selection of interactive multi model(IMM)model design,the transfer probability matrix of the traditional IMM algorithm is usually obtained based on prior information or artificially given,and is not adjusted in real time in combination with posterior information,so it will There is a large error.As the mobility gradually increases,the error will continue to increase,resulting in an unsatisfactory final tracking effect.This article makes the following improvements to the above problems:(1)For several nonlinear filtering algorithms commonly used in the target tracking process of GNSS system,the tracking accuracy and stability of these nonlinear filtering algorithms are compared and analyzed to provide an effective theoretical basis for subsequent research.Through MATLAB simulation,it is verified that the CKF algorithm has good stability and filtering accuracy compared with UKF and EKF algorithms,and the estimated performance is the best.(2)In view of the problem that the performance error of the current model or the subsequent measurement value anomaly occurs,the CKF algorithm will have a significant decline in the filtering performance effect or the dispersion of the filtering effect,an adaptive SCKF algorithm based on M estimation is proposed.This thesis replaces the traditional CKF algorithm with the traditional SRCKF algorithm.On the basis of the traditional SRCKF algorithm,a significantly improved Huber weight function is added to improve the performance processing framework of the traditional SRCKF algorithm and improve the robustness of the traditional SRCKF algorithm.Finally,the MAP evaluator,which is now more influential,is added.MAP is an effective processing.The method of noise statistical characteristics is a sub-optimal unbiased maximum post-testing evaluator based on Sage-Husa.This thesis uses the MAP estimator to optimize the performance characteristics of processing noise statistics,monitor and estimate unknown or inaccurate noise statistics in real time,and improve the tracking accuracy of the traditional SCKF algorithm.MATLAB simulation results show that the adaptive SCKF(M-ASRCKF)algorithm based on M estimation has strong anti-linearity,good anti-error interference ability,and the convergence speed and robustness results are excellent.(3)In the selection of filtering algorithm and IMM model design in the motion target tracking problem,the transfer probability matrix of the traditional IMM algorithm is usually obtained according to the prior information or given artificially,and there is no real-time adjustment combined with the posterior information.An IMMCKF algorithm based on GWO algorithm is proposed.The algorithm can use the interactive multi-model cubature Kalman filtering algorithm based on GWO algorithm for posterior information correction;add the GWO algorithm to the traditional IMM algorithm,use the improved GWO algorithm to identify various parameters in the Markov probability transfer matrix in real time,and improve the traditional interactive multi-model cubature Kalman filtering.The reduction of filtering accuracy and stability caused by the model conversion lag of the Kalman filter algorithm improves the switching accuracy between models and reduces the positioning error caused by the model conversion lag.After MATLAB simulation,the results show that the proposed IGWO-IMMCKF algorithm is more accurate in model conversion,and the tracking accuracy is also improved.The algorithm gets rid of the dependence on fixed model parameters,and has stronger tracking accuracy and robustness.