分布式自适应网格交互多模型JPDA算法  

Distributed Adaptive Grid Interacting Multiple Model Joint Probabilistic Data Association Algorithm

作　　者：许江湖[1] 刘忠[1] 党玲[2] 

机构地区：[1]海军工程大学电子工程学院,武汉430033 [2]海军大连舰艇学院,大连116018

出　　处：《武汉理工大学学报（交通科学与工程版）》2013年第2期292-296,共5页Journal of Wuhan University of Technology（Transportation Science & Engineering）

基　　金：中国博士后科学基金项目(批准号:20090461460);湖北省自然科学基金项目(批准号:2009CDB301)资助

摘　　要：将自适应网格交互多模型(AGIMM)算法和联合概率数据互联(JPDA)算法相结合,提出了用于多传感器多目标跟踪的分布式自适应网格交互多模型联合概率数据互联(DAGIMM-MSJP-DA)算法.该算法在航迹关联检验阶段结束后,融合中心采用模糊加权方法对不同传感器关于同一目标的各模型概率、各模型状态估计及其对应的估计协方差阵进行融合,从而得到关于目标的总体状态估计和相应的估计协方差阵,并将融合后的各模型概率反馈至各传感器以便进行统一的、更为精确的网格自适应调整.最后通过计算机仿真验证了该算法的有效性和正确性.Adaptive grid interacting multiple model （AGIMM）algorithm is combined with joint proba- bilistic data association（JPDA）, distributed adaptive grid interacting multiple model joint probabilistic data association （DAGIMM-MSJPDA） algorithm is proposed using for multisensor multitarget track- ing. The model probabilities, model-conditioned state estimation and its covariance are fused in fusion centre of this algorithm based on fuzzy weighted method after track association is finished, then the o- verall state estimation and its covariance can be obtained. The fused model probabilities is feed backed to all sonsors, so uniform and more accurate grid adaptation can be carried out. Validity and correct- ness of the algorithm is justified by computer simulation.

关 键 词：变结构多模型 多目标跟踪 多传感器 分布式 自适应网格 

分 类 号：TN957[电子电信—信号与信息处理]