基于AFKF的多传感器加权融合算法  

The Weighted Fusion Algorithm of Multi-sensor Information Based on Adaptive Fading Kalman Filter

作　　者：周少伟[1] 李洲洋[1] 雷少坤[1] 

机构地区：[1]西北工业大学机电学院,陕西西安710072

出　　处：《机械设计与制造工程》2013年第8期64-67,共4页Machine Design and Manufacturing Engineering

基　　金：国家自然科学基金资助项目(51205322);西北工业大学研究生创业种子基金资助项目(z2012049)

摘　　要：为实现对被测物体在匀速、匀加速以及变加速运动状态下的动态位置进行精确测量,提出一种基于自适应渐消卡尔曼滤波的多传感器加权融合算法,将各子传感器测量数据进行自适应渐消卡尔曼滤波,并由获得的均方误差阵实时为位置测量的估计值自适应分配权重,最终进行加权融合。通过算法的实例应用验证,与传统的MSIF-SKF算法相比,该算法具有更高的动态位置检测精度。In order to achieve accurate measurement of the measured object’s dynamic position in uniform,uniformly accelerated and varying accelerated motion state,it puts forward a weighted fusion algorithm of multi sensor information based on adaptive fading kalman filter.This algorithm makes the measurement data from each sub-sensor for adaptiving fading kalman filter,and then allocates the estimate of position measurement weights adaptively and real-time with the obtained mean square error matrix,at last,weighted fusion.Based on an example it validates the algorithm.Compared with the traditional MSIF-SKF algorithm,the WFMS-AFKF algorithm has a higher dynamic position detection accuracy.

关 键 词：多传感器 加权融合 自适应渐消因子 卡尔曼滤波 

分 类 号：TP391[自动化与计算机技术—计算机应用技术]