一种多传感器数据自适应空间分级融合算法  

Self-adaptive spatial stepped fusion algorithm for multisensor data

作　　者：贾俊[1] 李学仁[1] 卢虎[2] 杨瑞坤[2] 

机构地区：[1]空军工程大学工程学院,陕西西安710038 [2]空军工程大学理学院,陕西西安710038

出　　处：《传感器与微系统》2010年第10期119-121,134,共4页Transducer and Microsystem Technologies

基　　金：陕西省自然科学基础研究计划资助项目(SJ08S12)

摘　　要：当采用分布在不同空间位置上的多传感器观测值对测量噪声干扰下的参数进行融合估计时,被测量的空间分散性对融合结果影响较大。针对该问题,以自适应加权融合算法为基础,提出了自适应空间分级融合算法,并给出了误差分析和应用方法。该算法将融合过程分解为两次寻优,第1次是局部空间的自适应加权寻优,第2次是在全局空间内的融合寻优。计算机仿真结果表明:该算法在估计空间分布不均匀的被测量时优于自适应加权融合算法。When the survey data sampled by multiple sensors are adopted to estimate the parameter under the interference of measurement noise, the spatial variety of parameter to be measured influence the fusion result remarkably. In order to improve the measurement precision, a self-adaptive weighted spatial stepped fusion algorithm for muhisensor data fusion is presented based on self-adaptive weighted fusion algorithm, as well as error analysis and application method. The fusion process by this algorithm can be divided into partial space selfadaptive weighing optimum seeking and entire space fusion optimum seeking. The simulation results show that this algorithm is better than adaptive weighted fusion algorithm in estimation of spatial variety of measured value.

关 键 词：多传感器 数据融合 空间分布不均匀 空间分级融合 

分 类 号：TP274[自动化与计算机技术—检测技术与自动化装置]