多传感器和数据融合(一)

Multisensor and data fusion (Part 1)

文中介绍了多传感器集成及数据融合的概念、优点、结构、方法和应用，列举了传感器四种不同集成度的特点。数据融合把来自不同传感器的、或其它信息源的数据加以综合、相关、互联，以便提高定位和特征估计的精度。在数据融合过程中建模包括信号模型、噪声模型、变换器模型、数据变换模型以及融合模型。数据融合模型包括融合的方法和结构，文章介绍了集成式、分布式和混合式融合结构，并对它们进行了比较。此外，还介绍了国外一些数据融合的试验系统，商业软件和应用的例子。

The concept advantages architecture methods and application of multisensor integration and data fusion are described. Four degrees of multisensor integration are listed in the paper. Data fusion is the process of combining correlating associating data from multisensor and/or other information sources to achieve refined position identity estimation and assessment of situations. In the process of data fusion, the representation transform and processing of data must be modelled, which includes models of signal noise transducer data transform and fusion. Data fusion models include fusion method and architecture. The centralized distributed hybrid and hierarchical fusion architectures are described, and some comparison are made between them. In addition, some foreign data fusion systems commercial softwares and application samples for fusion are also presented.