环绕智能环境多模态数据融合技术研究

针对环绕智能环境中感知数据类型不同、格式多样问题,设计了基于XML的数据表示方法。其次,提出数据级和决策级相结合的数据融合架构,解决了环绕智能环境数据存在的异质、弱关联问题。最后,设计实现了基于加权自适应和模糊逻辑的数据融合算法,并应用于环绕智能环境监测,实验证明本技术可有效抑止传感器精度差异及噪声等因素导致的感知数据漂移问题,提高了系统决策的准确性。

Technology on Multi-Modality Data Fusion in Ambient Intelligence

According to the type and format diversity of Ambient Intelligent (AmI) environmental data, a XML-based representation for AmI data is firstly designed. Meanwhile, an architecture combining the data level and decision level data fusion is proposed to solve the AmI data problems on heterogeneity and weak association. Finally, a data fusion algorithm based on weighted adaptive and fuzzy logic is designed, and applied to AmI environmental monitoring. The experiment shows that this technology can reduce the error of data measurement and instability, and improve the decision accuracy of the system.