一种多红外传感器融合算法  

A Multi-Infrared-Sensor Fusion Algorithm

作　　者：刘颢[1] 李木华[2] 陈世友[1] 

机构地区：[1]武汉数字工程研究所,武汉430074 [2]黄冈供电公司,湖北黄冈438000

出　　处：《火力与指挥控制》2011年第11期120-124,共5页Fire Control & Command Control

基　　金：国防重点预研基金(51306030201);国防基金课题资助项目(9140A06040111CB39)

摘　　要：多红外传感器跟踪系统常常由于目标机动或传感器相对目标位置导致距离估计信息不可用。一种新的多红外传感器融合算法,采用一种改进的交互多模型滤波器,实现了对机动目标的三坐标跟踪。通过分析多红外跟踪系统的可观性,证明了系统的不可观定理;同时对算法进行了仿真实验,定理和仿真结果均表明算法有效。The range information estimated by Multi-Infrared-Sensor tracking system is often invalid, because of the target＇s maneuver and the sensor＇s position relative to the target.The new Multi-InfraredSensor fusion algorithm uses an improved Interactive Multi-Model（IMM） tracking algorithm based on Extended Kalman Filter（EKF）,to track the maneuvering target in 3-Dimension.A unobservable theorem of the system is proved by analysising the obervability of Multi-Infrared-Sensor tracking system and a simulation test on the algorithm is carried out.Both the theorem and the simulation results showed that the designed fusion algorithm is valid.

关 键 词：多红外传感器 机动目标跟踪 可观性 局部可观性 交互多模型 

分 类 号：TP274[自动化与计算机技术—检测技术与自动化装置]