多传感器数据融合的主成分方法研究  

Study on principle component method for multi-sensor data fusion

作　　者：董九英[1] 

机构地区：[1]江西财经大学信息管理学院,南昌330013

出　　处：《计算机工程与应用》2009年第33期111-113,共3页Computer Engineering and Applications

基　　金：国家自然科学基金No.10626029;江西省自然科学基金No.0611082;江西省教育厅科技项目(No.GJJ08350)~~

摘　　要：针对多个传感器对某一特性指标进行测量实验的数据融合问题,提出了一种基于主成分分析的融合方法。该方法把各传感器的测量数据作为一变量,定义总体的各主成分,利用测量值与主成分的复相关关系,给出了各传感器的综合支持程度和数据融合公式。应用实例验证了方法的有效性和精确性。Due to data fusion of multi-sensor experiment on some characteristic index,a new fusion method is proposed based on the principle component analysis.The method views the measured data of every sensor as a variant.After defining each principle components for the collectivity,the synthesis support degrees of all sensors are given according to the compound relationship between the measured value and the principle component.The formula of data fusion is obtained.The applied example proves that the method is both effective and accurate.

关 键 词：多传感器 数据融合 主成分分析 综合支持度 

分 类 号：TP212[自动化与计算机技术—检测技术与自动化装置]