摘要
研究了一类非线性时变动态系统的状态估计问题,在不同传感器以不同采样率异步对同一目标进行观测时,提出了一种有效的数据融合估计算法.通过建立多尺度模型,将异步多速率系统形式转化为同步多速率系统;在每一步分别进行状态的预测和更新.在状态和观测预测时,采用强跟踪滤波(STF)算法;在状态更新时,采用有反馈分布式结构,顺序的利用每一个传感器的观测信息去更新状态的估计;从而基于给定的非线性系统模型,得到融合所有异步、多速率传感器观测信息的状态估计结果.该方法不需要对状态或观测进行扩维,计算量适当,从而保证了算法的实时性.仿真结果验证了算法的有效性.
A kind of time-vary nonlinear dynamic system is studied in this paper.An effective data fusion state estimation algorithm is presented in time of multiple sensors observing a single target with different sampling rates asynchronously.The asynchronous multirate system is transformed to synchronous multirate system by use of the established multiscale models.In each step,to get the state estimate,state prediction is followed by state update.In state and measurements prediction step,strong tracking filter （STF） is used. While, in state update step, distributed structure with feedback is used, and the fused state estimate is obtained by se- quentially use of the measurements observed by different sensors. The augmentation of state or measurement dimensions are avoided by use of the presented method, and the real-time property of the algorithm is guaranteed. Simulation results show the effectiveness of the proposed algorithm.
出处
《电子学报》
EI
CAS
CSCD
北大核心
2009年第12期2735-2740,共6页
Acta Electronica Sinica
基金
国家自然科学基金(No.60773044)
教育部创新团队支持计划资助项目
关键词
数据融合
非线性系统
异步
多速率
强跟踪滤波
data fusion
nonlinear system
asynchronous
multirate
strong tracking filter
