机动目标当前统计模型模糊自适应算法.

A fuzzy adaptive algorithm is proposed for the imperfections of tracking a maneuvering target using conventional algorithm based on current statistical model. Maneuvering frequency is adjusted in real time by fuzzy reasoning according to the normalized residual and its change rate. Acceleration variance is depicted using residual power function, and the mean value of current acceleration is updated by the de- viation between the estimated and predicted values of acceleration. On this basis, the weight of proposed algorithm is revised by Gauss membership function and strong tracking algorithm. Fuzzy adaptive algorithm is not restricted by maneuvering frequency given manually and extreme value of maximum acceleration, which is suitable for the different ranges and degrees of maneuvering. The performance of the proposed algorithm is tested by tracking three typical maneuvering targets, such as step maneuvering, circular maneuvering, and Jerk maneuvering. Simulated results show the tracking range is expanded using the pro- posed algorithm compared with the conventional tracking algorithm based on current model and the adaptive algorithm based on Jerk model. The proposed algorithm has good steady-state and transient characteristics, and its tracking accuracy and convergence rate are superior to those of the other two algorithms. [ABSTRACT FROM AUTHOR]