摘要
以异步多速率传感器信息融合理论和同步单速率传感器不完全观测理论为基础,提出了一种适用于时变线性系统的异步多速率传感器不完全观测的信息融合算法.通过原理分析和数学推导,将异步多速率传感器动态系统建模为同步同速率系统,进而利用改进的卡尔曼滤波方法进行状态估计,利用联邦分布式数据融合方法进行信息融合,获得基于所有观测信息的最优估计.理论分析和仿真结果均表明该算法的融合效果优于任一单传感器卡尔曼滤波的效果.
An information fusion algorithm based on a class of asynchronous multi-rate and multi-sensor system was presented,where measurements from each sensor may be randomly missed with a certain probability.The fused result was obtained in virtue of information fusion theory of asynchronous multi-rate sensors and the approach for fusing missing measurements based on synchronous single-rate sensor system.According to theoretical analysis,we built the synchronized single-rate system model for dynamic system of asynchronous multi-rate sensors.So the approach could be adopted in time varying system with asynchronous multi-rate sensors.By using the improved Kalman filter for state estimation and federal filter for information fusion,we gain the optimal estimation based on all the observations.Theoretical analysis and simulation results show the efficiency of the proposed algorithm comparing with others.
出处
《华中科技大学学报（自然科学版）》
EI
CAS
CSCD
北大核心
2009年第S1期271-274,共4页
Journal of Huazhong University of Science and Technology(Natural Science Edition)
基金
国家高技术研究发展计划资助项目(2006AA12A103)
国家杰出青年科学基金资助项目(60625102)
国家自然科学基金重点资助项目(60532030)
关键词
信息融合
线性系统
时变系统
传感器
异步多速率
不完全观测
卡尔曼滤波
information fusion
linear system
time varying system
sensor
asynchronous multi-rate
missing measurements
Kalman filtering
