摘要
针对密集杂波环境下对多目标跟踪的精度低、实时性不强的问题,提出了密集杂波下模糊聚类数据关联多目标跟踪算法。该算法利用模糊聚类,得到不同观测量相对目标的隶属度作为模糊关联概率,通过分析公共观测对目标的影响,引入远近距下的公共观测影响因子重建模糊关联概率矩阵;然后结合模糊关联概率与卡尔曼滤波,对不同观测量得到的状态估计加权融合,从而对每个目标进行单独跟踪,实现目标的状态更新。仿真结果表明,杂波密集环境下该算法在能够保证多目标跟踪实时性的同时引入远近距下公共影响因子对不同观测量的状态估计进行加权,保证了目标跟踪的精确性。
Aiming at the problem of low precision,large amount of calculation and low real-time performance for multi-target tracking in dense clutter environment,a multi-target tracking algorithm based on fuzzy clustering data under dense clutter was proposed.The algorithm used fuzzy clustering to obtain the degree of membership of different target relative observations as the fuzzy association probability.Through the analysis of the influence of public observations on the target,the public observation impact factors under long-distance distance to reconstruct fuzzy associated probability matrices was introduced.Then,combining the fuzzy association probability and the Kalman filter,the state estimations obtained from different observations were weighted and fused,so that each target was separately tracked and the state of the target was updated.The simulation results showed that the algorithm could reduce the real-time performance of multi-target tracking while introducing the long-distance distance common impact factor to weight the state estimation of different observations,which ensured the accuracy of target tracking.
作者
康旭超
何广军
陈峰
何其芳
KANG Xuchao;HE Guangjun;CHEN Feng;HE Qifang(Air and Missile Defense College,Air Force Engineering University,Xi'an 710051,China)
出处
《探测与控制学报》
CSCD
北大核心
2019年第4期56-61,65,共7页
Journal of Detection & Control
基金
国家自然科学基金项目资助(61703424)
关键词
模糊聚类
多目标跟踪
数据关联
卡尔曼滤波
fuzzy clustering
multi-target tracking
data association
Kalman filter