基于容积卡尔曼滤波的异类多传感器一致性融合算法  

Heterogeneous Multi-Sensor Consistency Fusion Algorithm Based on Cubature Kalman Filter

作　　者：金长江[1] 张兵[1] 曹祥杰[1] 

机构地区：[1]中国电子科技集团公司第二十七研究所,郑州450047

出　　处：《传感技术学报》2016年第7期1006-1010,共5页Chinese Journal of Sensors and Actuators

基　　金：863计划基金项目(2015AA8098089A)

摘　　要：针对单一光频传感器获取目标特征信息存在的不一致性,提出一种基于容积卡尔曼滤波的异类多传感器一致性融合方法。首先,从原理上分析了激光、红外与雷达三类传感器量测信息的特征及其存在的差异,进而在容积卡尔曼滤波框架下,针对雷达、红外和激光探测等组成的典型目标侦测系统,结合一致性融合策略,通过对目标距离和方位信息融合处理改善目标状态估计精度。仿真结果表明:相对于传统的单传感器滤波方法,所提出的融合方法和策略具有较好的滤波性能。Aiming at the inconsistency of target feature information obtained form single optical frequency sensor ,a novel heterogeneous multi-sensor consistency fusion algorithm based on cubature Kalman filter is proposed. Firstly, the characteristics and differences of measurement information from three kinds of typical sensors including laser , infrared and radar are discussed in theoretical. Secondly,for the typical target reconnaissance system under the framework of Cubature Kalman filter,combining with consistency fusion strategy,the target state estimation filtering is improved by fusion processing of distance and azimuth information. Finally,the target state information after fusing was acquired. The simulation results show that the proposed algorithm can obtain better filtering performance relative to the traditional single sensor filtering method.

关 键 词：非线性滤波 异类多传感器 容积卡尔曼滤波 一致性融合 

分 类 号：TP393[自动化与计算机技术—计算机应用技术]