多传感器集中式增量卡尔曼滤波融合算法  

Multi-sensor centralized incremental Kalman filtering fusion algorithm

作　　者：马丽丽[1] 张曼[1] 陈金广[1] 

机构地区：[1]西安工程大学计算机科学学院,西安710048

出　　处：《计算机工程与应用》2014年第11期229-232,255,共5页Computer Engineering and Applications

基　　金：国家自然科学基金(No.61201118);中国博士后科学基金项目(No.2013M532020);陕西省教育厅科研计划项目(No.12JK0529)

摘　　要：在单个传感器的状态估计系统中,标准的增量卡尔曼滤波方法可以有效消除量测系统误差。对于多传感器情况,标准算法失效。针对该问题,提出了多传感器集中式增量卡尔曼滤波融合算法,即:增量卡尔曼滤波的扩维融合算法和增量卡尔曼滤波的序贯融合算法。在标准增量卡尔曼滤波算法的基础上,结合扩维融合和序贯融合的思想来实现多传感器数据的融合。实验结果表明,当存在量测系统误差时,提出的集中式融合算法与传统的集中式融合算法相比,提高了滤波精度,并且能够成功地消除量测系统误差。In the state estimation system of single sensor, standard incremental Kalman filter can eliminate measurement system error effectively. For the system of multi-sensor, standard algorithm does not work. To the problem, the paper presents multi-sensor centralized incremental Kalman filtering fusion algorithms, i.e., augmented fusion algorithm with incre-mental Kalman filter and sequential fusion algorithm with incremental Kalman filter. Based on the standard incremental Kalman filter, multi-sensor data fusion is implemented by using the ideas of augmented fusion and sequential fusion. Experi-mental results show that the fusion accuracy of the proposed centralized fusion algorithms is better than that of traditional centralized algorithms. Furthermore, the new fusion algorithms can eliminate the measurement system error.

关 键 词：多传感器数据融合 增量卡尔曼滤波 扩维融合 序贯融合 量测系统误差 

分 类 号：TP391[自动化与计算机技术—计算机应用技术]