杂波环境下雷达组网的多目标聚类融合跟踪

多传感器多目标跟踪是信息融合技术在目标跟踪领域的应用范例,数据关联是其中的关键技术之一.对于杂波环境下的组网雷达多目标跟踪,讨论了粗、精关联相结合的数据关联方法.先用基于跟踪门限算法进行粗关联,排除部分杂波,再用模糊C-均值算法模糊聚类来实现关联.通过把多传感器跟踪问题转化为多个单传感器跟踪问题,更有效地实现关联,最后融合量测,滤波后得到目标的状态估计.用该算法对目标进行蒙特卡罗仿真,其比改进前的模糊C-均值关联算法和最近邻域算法在杂波环境下更能有效实现数据关联.

Clustering based multi-target tracking of radar network in clutter

Multi-sensor multi-target tracking is a paradigm of information fusion technique dealing with target tracking problems.Data association is one key aspect.To improve multi-target tracking of a radar network in cluttered environment,a data association algorithm was proposed which includes two parts: coarse and precision correlation.First,some interference was eliminated during coarse correlation based on the gate algorithm;then the remaining measurements were fuzzy clustered using FCM(Fuzzy C-Means) algorithm.This was done by decomposing the multi-sensor multi-target problem into data association problems of several single sensor multi-targets.Finally,a state estimation of targets was gained after measure-to-measure fusion.Monte-Carlo simulation results show the proposed algorithm is better than previous modified algorithms and NN algorithms for data association in clutter.