基于UKF的变采样率多异质传感器异步数据融合  

Variable Sampling Rate Multiple Dissimilar Sensors Asynchronous Data Fusion Based on UKF

作　　者：周样晶[1] 江晶[2] 赵岐诚[3] 陈于平[4] 

机构地区：[1]空军雷达学院研究生管理大队,武汉430019 [2]空军雷达学院信息与指挥自动化系,武汉430019 [3]中电集团第十四研究所,南京210013 [4]解放军通信指挥学院通信与指挥系统教研室,武汉430019

出　　处：《传感技术学报》2008年第5期795-800,共6页Chinese Journal of Sensors and Actuators

基　　金：国防“十一五”预研项目资助(51307020204)

摘　　要：针对异质传感器数据融合能够实现信息互补,改善目标跟踪精度,提出了一种多异质传感器在变采样率下的异步量测融合算法,即首先将多传感器数据组合成类似于单传感器数据的异步数据处理方法,进行点迹合成,再将合成后的虚拟量测对当前时刻的目标状态进行更新.变采样率跟踪是基于网络或栅格多传感器异步融合跟踪的基础,通过引入时戳的概念给出了基于UKF(Unscented Kalman Filter)的具体融合算法,最后通过仿真验证了该算法的有效性。According to dissimilar sensor asynchronous data fusion can achieve information implementarity and enhance target track precision, a dissimilar multisensor asynchronous data fusion arithmetic based on variable sampling rate is presensed. Firstly, using asynchronous data dealing way of composing multisensor data to single-sensor data, by which point track is composed; secondly, current target state is updated using composed measurement. Variable sampling rate tracking is based on nets or grids multisensor asynchronous fusion tracking. It gives embody fusion arithmetic based on UKF by introducing on time stample. Finnaly, simulations result validates validity of the algorithm.

关 键 词：异质 异步 融合 变采样率 时戳 UKF 

分 类 号：TP212[自动化与计算机技术—检测技术与自动化装置] TP274[自动化与计算机技术—控制科学与工程]