分段跟踪识别器的多特征加权指标设计

分段跟踪识别器（STI）解决的是机动目标相邻运动模式切换时刻判定和运动模式参数辨识互相耦合的问题，这种非贝叶斯的方法不需要先验给定目标的运动模型集与模型之间的转移概率。STI在设计性能指标时主要存在以下两个问题：一是仅考虑目标位置与航向角连续性，特征量不完备；二是各连续性指标权重先验给定，缺乏设计准则。为此，提出了STI的多特征加权指标设计方法。首先，引入了速度连续性指标，改善了目标运动模式提取时特征量的完备性；然后，基于各指标的协方差矩阵对其权重分别进行了设计，消除了各指标量纲的影响。仿真结果表明，与STI相比，新提出的方法运行速度快，目标运动模式误切换次数与切换时延明显下降，目标位置与角速度估计精度也有所提升。

Multi-feature weighted index design of segmentation track identifier

Segmentation track identifier(STI) solves the problem of mutual coupling between the determination of the switching moment of the adjacent motion mode and the identification of the motion mode parameters for maneuvering target.This non-Bayesian approach does not require a priori given set of motion models and the transition probability between models of the target.It is found that STI has the following problems in designing performance indicators:only the continuity of position and heading angle of the target is considered,and the feature quantity is incomplete;the weights of the continuity indicators are given a priori and there is a lack of design criteria.To this end,a multi-feature weighted index design method for STI is proposed.Firstly,the speed continuity index is introduced to improve the completeness of the feature quantity in extracting the target motion pattern.Then,the weights of the index are designed based on the covariance matrix of the corresponding index,which eliminates the influence of the index dimension.Simulation results show that the proposed method has faster running speed,better target motion mode switching accuracy and higher accuracy of target state estimation compared with the STI method.