基于Multi-Agent技术的三层信息融合系统研究

针对传统传感器网络管理复杂,系统信息融合智能化不高、精度低和模式单一、结构不清晰等不足,首先分析了Multi-Agent技术、传感器网络技术以及信息融合技术的独特优势,然后采用计算机网络分层结构思想和基于人工智能本体的知识表达理念,在信息融合过程中采用改进的SVM分类方法,构建了一种基于Multi-Agent技术的多传感器三层信息融合系统并对其具体融合过程进行了分析。最后对分类过程用MATLAB进行了分析。实验结果表明:系统分类精度较高,一定程度上不仅明显弥补了传统传感器的诸多不足,而且为后期决策提供了较为精准的目标参数。

Three-tier information fusion system research based on Multi-Agent technology

For some shortcomings of the traditional sensor networks,such as:complex network management,low degree of system integration intelligent,low –precision,single model,the structure is not clear,etc.,The unique advantages of the Multi-Agent technology,sensor networks technology,information fusion technology are first analyzed,and then the computer network hierarchy ideas and knowledge representation concept based on artificial intelligence ontology and a improved method of SVM classification for information fusion to construct a multi-sensor three-tier information integration system based on Multi-Agent technology are used and then its specific integration process is analyzed.At last,the classification process using MATLAB is analyzed.Experimental results show that:the system has high classification precison and to a certain extent not only obviously make up for the shortcomings of conventional sensor but also provides some more accurate targent parameters for late decision-making.