关键词: 航迹融合, 当前 统计模型, 卡尔曼滤波, 新息差序列, 方差自适应

Abstract:

In the case of the strong maneuver such as turning and accelerating, the tracking fusion accuracy of the traditional current statistical model goes sharply declined. To solve this problem, we proposed an improved model by extracting the information of the residual error sequence and measuring the variance sequence. The new model we proposed can adaptive adjust to the maneuvering frequency and maximum acceleration by the detection of the maneuvering mode and the addition of the covariance adaptive factor in the kalman filtering framework. Through the improvement, the algorithm can not only keep the good performance of general maneuvering target tracking, but also improve the matching degree between current statistical model and the real motion pattern in the situation of target maneuvering state transition, which makes the fusion accuracy of the track fusion system greatly improved when the target maneuvering state is abrupt. Finally, Monte Carlo simulation is used to verify the proposed method.

Key words: track fusion;current statistical model;kalman filter;innovation sequence;variance adaptive