%0 Journal Article
%T 多模型GGIW-GLMB算法跟踪机动群目标<br>Maneuvering group target tracking with multi-model GGIW-GLMB algorithm
%A 甘林海
%A 刘进忙
%A 王刚
%A 李松
%J 北京航空航天大学学报
%D 2018
%R 10.13700/j.bh.1001-5965.2018.0053
%X 摘要 针对多个机动群目标跟踪问题，提出了一种多模型伽马高斯逆威夏特-广义标签多贝努利（MM-GGIW-GLMB）算法。采用多模型算法对群目标进行运动建模，利用最适高斯（BFG）近似在预测阶段对多模型进行融合，减小了多模型算法的运算量，为进一步提高算法在目标机动阶段的跟踪性能，引入强跟踪滤波器（STF）对BFG算法得到的预测状态协方差进行修正。利用最优次模式分配（OSPA）距离及其一倍标准差和航迹标签正确率衡量算法对机动群目标的跟踪性能。仿真结果表明，本文算法能够提升对机动群目标的跟踪精度和稳定性。<br>Abstract：An multi-model Gamma Gaussian inverse Wishart-generalized labeled multi-Bernoulli (MM-GGIW-GLMB) algorithm is proposed for multiple maneuvering group target tracking. A multi-model approach is introduced for kinematic modeling, and best fitting Gauss (BFG) approximation is used to fuse the multiple models in the prediction stage, which subsequently ease the computational burden of multi-model approach. For a further performance improvement for target maneuvering, strong tracking filter (STF) is introduced to correct the predicted covariance calculated by BFG. The optimal sub-pattern assignment (OSPA) metric and its one standard deviation and labeling correctness are used to measure the maneuvering group target tracking performance of the algorithm. The simulation results indicate that the proposed algorithm can improve the performance of maneuvering group target tracking in accuracy and stability.
%K 广义标签多贝努利(GLMB)
%K 伽马高斯逆威夏特(GGIW)
%K 最适高斯(BFG)近似
%K 强跟踪滤波器(STF)
%K 群目标跟踪&lt
%K br&gt
%K generalized labeled multi-Bernoulli (GLMB)
%K Gamma Gaussian inverse Wishart (GGIW)
%K best fitting Gauss (BFG) approximation
%K strong tracking filter (STF)
%K group target tracking
%U http://bhxb.buaa.edu.cn/CN/abstract/abstract14618.shtml