基于UKF的异类传感器集中式融合算法  

UKF-based Centralized Fusion Algorithm for Inhomogeneous Sensors

作　　者：陈嘉鸿[1] 张新曼[1] 张忠华[2] 韩九强[1] 

机构地区：[1]西安交通大学电信学院,西安710049 [2]中国卫星海上测控部,江苏江阴214431

出　　处：《航天控制》2008年第5期3-6,17,共5页Aerospace Control

基　　金：国家自然科学基金资助项目(60502021)

摘　　要：综合考虑航天器跟踪测量中速度和精度的要求,对异类传感器集中式融合问题进行了研究,提出了基于无迹卡尔曼滤波(UKF,Unscented Kalman Filter)和简化的分类数据压缩技术的非线性系统实时集中式融合算法。仿真表明新的算法融合性能优于基于扩展卡尔曼滤波(EKF,Extended Kalman Filter)和扩维技术的并行集中式融合算法。Considering of precision and speed in spacecrafis tracking, the centralized fusion algorithms for inhomogeneous sensors are studied. A new real-time algorithm based on unscented Kalman filter （UKF） and simplified classifying data - compression （ DC ） technology is proposed for nonlinear multi-sensor data fusion. In view of real time precision, the new algorithm is suitable for multi-sensor single-target tracking, especially for the local center of space measuring. The simulation results show that the performance of the new algorithm is superior to the extended Kalman filter （EKF） and expanded-dimension （ED） technology based parallel centralized fusion algorithms.

关 键 词：无迹卡尔曼滤波 集中式融合 异类传感器 数据压缩 

分 类 号：V557.1[航空宇航科学与技术—人机与环境工程]