作　　者： ; ;

机构地区： 深圳大学信息工程学院ATR国防科技重点实验室

出　　处： 《信号处理》 2011年第9期1301-1305,共5页

摘　　要： 多目标跟踪中的数据关联一直是信息融合领域的难点和热点问题,针对杂波环境下多目标跟踪中的数据关联问题,提出了一种基于模糊推理的JPDAF新方法。该方法中,首先详细分析了杂波环境多目标观测数据的特点,定义了多目标环境下的标准化新息变量及新息的一阶微分变量;然后将其作为模糊推理的两个输入变量,通过设计合适的模糊隶属度函数和模糊推理规则,自适应计算目标观测的关联概率来代替传统联合概率数据关联滤波器(JPDAF)中的关联概率,实现对多个目标的有效跟踪。实验结果表明,提出方法的目标跟踪性能要好于传统的JPDAF和Fitzgerald's方法,在实时性方面,提出方法也要远好于传统的JPDAF方法,接近Fitzgerald's方法,能够有效对多目标进行关联跟踪。 The problem of data association is very important to multiple target tracking in informantion fusion field.To the data association problem of the multiple target tracking in clutter environment,a new joint probabilistic data association filter（JPDAF） method based on fuzzy logic inference was proposed.Firstly,the characteristics of the measurement of multiple targets were analyzed.Secondly, the rules of the proposed algorithm are expressed in terms of two input variables and one output variables.The input variables are defined in terms of the prediction error and change of error.At the same time,in order to adapt computing the asscoaition probabilities of targets and measurements,the fuzzy inference rules were designed according to the advices of experts.Thirdly,the association probabilities of the joint probabilistic data association filter is replaced by the adaptive membership degrees that is computed through fuzzy inference system.Finally,the simulation results show that the performance of target tracking of the proposed algorithm is higher than the JPDAF and Fitzgerald＇ s method,the runtime statistics of each algorithm is show that the proposed algorithm is faster than the JPDAF, and inferior to the Fitzgerald＇ s method.

关 键 词： 多目标跟踪 模糊推理 数据关联

领　　域： [电子电信] [电子电信]