An improved adaptive maneuvering target tracking algorithm

During the process of maneuvering target tracking, the difference between observation value and extrapolated value is processed as noise or maneuvering acceleration according to the range of the difference. However the maneuvering target could not be tracked while the difference is large. In this paper, a new adaptive α-β algorithm is brought out. It is that two extrapolated values are considered, while the target is being tracked. They are normal extrapolated value and maneuvering extrapolated value. In the end of the paper, the concerning experiments are introduced. The result is that the maneuvering target can be continuous tracked. The performance of the new algorithm is also compared with that of Kalman filter.