次最优概率和集中式MSJPDA算法  

Sub-optimal MSJPDA algorithm based on probability weight

作　　者：张琦[1,2] 张晶炜[1,2] 熊伟[2] 何友[2] 

机构地区：[1]海军92301部队,北京100020 [2]海军航空工程学院信息融合技术研究所,烟台264001

出　　处：《哈尔滨工业大学学报》2009年第5期188-191,共4页Journal of Harbin Institute of Technology

基　　金：国家自然科学基金资助项目(60172033);全国优秀博士论文作者专项资金资助项目(200036)

摘　　要：针对集中式多传感器联合概率数据互联(MSJPDA)并行算法经常出现滤波发散的现象,以及集中式MSJPDA算法实现复杂且运算量大的问题,对并行结构的MSJPDA算法进行了合理的修正,修正后的算法在滤波时运用概率数据互联的思想对各传感器的修正量进行概率加权,将次最优联合概率数据互联算法引入到集中式MSJPDA算法,简化后算法在性能接近的情况下有效地减少了运算量.仿真比较与分析结果表明,本文算法的综合性能更优越.To solve the problems that the parallel implementation outcome of centralized Muhisnesor Joint Probabilistic Data Association （MSJPDA） algorithm filter is often emanative in simulations, and the MSJPDA algorithm has a high computation burden, a modified parallel centralized MSJPDA algorithm is presented. The modified algorithm provides a coefficient for correction of each sensor using probabilistic data association technique in state estimation. Then, suboptimal JPDA technique is applied to the MSJPDA algorithm and the computational cost of simplified algorithm is reduced. Some typical simulations are performed and the results show that tracking performance of the algorithm is much better than that of MSJPDA existed.

关 键 词：集中式 多传感器 多目标跟踪 数据互联 

分 类 号：TN95[电子电信—信号与信息处理]