Hypersonic Glide Target Tracking Based on Improved Square Root UKF

This paper mainly proposes an improved square root UKF filtering algorithm (ISR-UKF) to solve the problem of near space hypersonic glide vehicle (HRGV) tracking. The algorithm firstly adopts the general algorithm of square root UKF (SR-UKF) and then uses the spherical unscented transform to calculate weight coefficients and sigma points, and introduces singular value decomposition method to decompose the covariance matrix into square root form for iterative update algorithm. At the same time, according to the characteristics of such targets, the target state equation is established on the basis of the aerodynamic model. Furthermore, the tracking effect of the balance glide trajectory and jump glide trajectory of the HRGV target is simulated and analyzed by using this algorithm with the square root UKF algorithm via singular value decomposition (SVDSR-UKF) and the SR-UKF. The results indicate that the ISR-UKF algorithm has higher filtering accuracy, faster convergence speed and better stability, and can achieve good tracking of HRGV targets.