Abstract: Based on the analysis of the applicabilities and limitations of traditional current statistical models, an improved current statistical model adaptive filtering algorithm is proposed to address the problem of low tracking accuracy caused by the need to pre-set fixed values for acceleration limit and maneuvering frequency in the current statistical model algorithm, which cannot adaptively adjust algorithm parameters during maneuvering target tracking. During the maneuvering target tracking, the algorithm adaptively adjusts the variance of acceleration and maneuvering frequency through the relationship between the difference in position estimation and acceleration disturbance, as well as the mean of the current acceleration estimation value, and then adaptively adjusts the variance of process noise. The results of Monte Carlo simulation experiments show that the improved algorithm can continuously and stably track targets, and has better performance than traditional algorithms in tracking weak and strong maneuvering targets, which improves the tracking accuracy.