Adaptive Tracking Algorithm Research For Hypersonic Target

The application of maneuvering target tracking technology is extremely widely in military field,in the joint efforts of many experts and scholars in recent years,the key algorithm research of this technology had made great progress and provide theoretical basis for precision strike and national air defense.But with the rapidly development of High-tech weaponry,and anti-tracking technology is developing constantly,especially when the concept of new strategic weapon-Near Space Hypersonic Vehicle was proposed,realized and developed,lead the existing air defense and antimissile system get serious challenged.First of all,introducing research background and research situation of the subject,focusing on adaptive filtering algorithm and model algorithm of maneuvering target tracking,particularly introducing several commonly used nonlinear target tracking filter algorithm such as EKF、UKF、CKF and QKF,and Interacting Multiple Model Algorithm.Analyzing the advantages and weaknesses of these algorithms under different environment,comparing the capability of these algorithms with simulation,and do analysis and summary to the experimental results.Secondly,the CQKF algorithm which has a better performance is introduced to track Near Space Hypersonic Vehicle.CQKF algorithm is a more generalized form of CKF algorithm,not only could overcome the problem of “curse of dimensionality”,but also the algorithm precision will increase with the increasing order of Gaussian-Laguerre integral points.The simulation result under the background of Near Space Hypersonic Vehicle shows that three-order CQKF algorithm has a better precision than CKF algorithm,followed by UKF and EKF.Then,when tracking the Near Space Hypersonic Vehicle with CQKF algorithm,the tracking precision may reduce or even losing the target caused by filtering divergence,aiming at this problem,an adaptive ST-CQKF algorithm is proposed.In the covariance matrix updating step of CQKF,drawing lessons from STF algorithm,the fading factor is introduced to real-time correct the prediction error covariance matrix and force the residual orthogonality,the stability and robustness of the algorithm is improved.The simulation result shows that ST-CQKF algorithm is more adaptable and a better stability.Finally,when STF make corrections to CQKF algorithm,a great difference in numerical of output vectors may lead asymmetric information in residuals,and then influence the computational efficiency of the algorithm,reduce the stability and accuracy,aiming at this problem,a residual normalization RNST-CQKF algorithm is proposed.In the calculation of fading factor,the filtering residuals are normalized to balance the sensitivity of different orders of magnitude information,thus improve the stability and precision of the algorithm.The simulation shows that RNST-CQKF algorithm not only improve the tracking accuracy but also reduce the running time of the algorithm and more computing resources are saved.