A basic current statistical model can only be applied to maneuvering target with high acceleration because its upper and lower limits of target acceleration is fixed and invariable. So a basic current statistical model and an adaptive filter algorithm has a good performance on maneuvering target with high acceleration but a poor performance on maneuvering target with low acceleration. In order to solve this problem, a novel fuzzy membership function is presented which uses the current acceleration of maneuvering target to adjust the upper and lower limits of target acceleration adaptively. This new algorithm makes sure that the current statistical model can be applied to maneuvering target with any acceleration. Finally, track prediction simulation is did which uses this new algorithm and CSAF algorithm to predict the track of maneuvering target. Simulation results show that, the prediction performance of this new algorithm is much better than CSAF algorithm on maneuvering targets both with low and high acceleration.