统计线性化无序量测更新算法  

Out-of-Sequence-Measurement Algorithm Based on Statistically Linearized Method

作　　者：王炜[1] 李丹[2] 黄心汉[3] 

机构地区：[1]海军工程大学理学院,武汉430033 [2]武汉理工大学理学院,武汉430070 [3]华中科技大学,武汉430074

出　　处：《火力与指挥控制》2014年第9期76-81,共6页Fire Control & Command Control

摘　　要：在实际的多传感器系统中,由于各种传感器具有不同的采样率,预处理时间,以及数据通信延迟,因而会出现多个传感器量测不同步到达融合中心的现象。进一步地,当较早时刻产生的量测在较晚时刻产生的量测之后到达融合中心时,无序量测问题就出现了。针对离散时间非线性系统,提出基于统计线性化固定点平滑器的最佳统计线性化无序量测算法,它可以处理单步和两步延迟(甚至多步)无序量测,且能达到与重新滤波法相同的估计精度。In a practical multi-sensor system,various sensors have different sampling rates, pretreatment time,and data communication delay,thus resulting in the phenomenon that a lot of sensor measurements don’t arrive at the fusion center synchronously. Further,when the measurements generated at an earlier moment arrive at the fusion center after the measurements generated at a later time,the OOSM issue occurs. For the discrete -time nonlinear systems,the optimal statistically linearized OOSM algorithm based on the statistically linearized fixed-point smoother which can deal with single-step delay OOSMs and double-step delay OOSMs and have the same accuracy with the re-filtering algorithm.

关 键 词：目标跟踪 滤波估计 无序量测 

分 类 号：TP273[自动化与计算机技术—检测技术与自动化装置]