An improved target tracking singer filter algorithm

Considering the low performance of accuracy and convergence of the traditional Singer algorithms for maneuvering detection, this paper proposed an improved Singer algorithm which deals with the adjustments of the matrix of the process noise covariance and changes of the filtering gain according to the average attenuation of innovation and the filtering value of acceleration. The new algorithm can reduce the RMSE of positions, and has the advantages of better filtering accuracy of velocity and acceleration. And the feasibility of the algorithm is proved by MATLAB simulations.