基于IEPF序贯融合的多传感器管理方法  

Research on Multi-sensor Management in Sequential Fusion based on Iterative Extended Kalman Particle Filter

作　　者：黄俊[1] 于雷[2] 李骞[2] 张斌[2] 

机构地区：[1]空军装备研究院总体所,北京100076 [2]空军工程大学工程学院,西安710038

出　　处：《火力与指挥控制》2010年第8期71-74,共4页Fire Control & Command Control

基　　金：空军武器装备科研基金(KJ06085);军队院校优秀博士学位论文创新基金资助项目(BC07005)

摘　　要：针对主被动传感器信息融合的特点,提出了一种在IEPF序贯融合基础上进行传感器管理的方法。工程常采用的传统扩展卡尔曼滤波融合算法滤波精度不高,因此结合采用了粒子滤波方法。先对主被动传感器采用迭代扩展卡尔曼滤波集中式序贯融合,利用融合后的信息进行粒子更新,提高滤波精度,在此基础上采用分辨力增益的方法对传感器进行管理。仿真结果表明该方法能够提高对目标的跟踪精度,增强多传感器对环境变化的适应能力。According to data Fusion of active and passive sensors,a sensor management algorithm was addressed based on centralized sequential filtering in this paper based on IEPF.Particle filtering was incorporated for improving the performance of IEKF,which is very effective in engineering.Combining active sensors with passive sensors by sequential filtering,information of filtering result was used for particle update.Discrimination gain in the IEPF were used for sensor management.Simulation results show that the novel method can significantly improve the multi-target tracking performance and adaptive capability of multi-sensor to change of circumstance.

关 键 词：迭代扩展卡尔曼粒子滤波 序贯融合 分辨力增益 传感器管理 

分 类 号：TP273[自动化与计算机技术—检测技术与自动化装置]