摘要
针对分布式多传感器环境下的目标识别问题,提出了一种基于加权D-S证据理论组合规则的决策融合方法。分析了多传感器目标识别系统的信息模型,指出传感器的决策可信度由其被支持度及与目标间的距离确定。将该可信度体现为加权D-S证据理论组合规则中的证据权值,综合考虑传感器支持度及其与目标距离,给出了权值确定方法。仿真实验证明方法提高了融合效率,可较快完成识别任务。
Dempster-Shafer （D-S） evidence theory has been widely used in data fusion systems. A method based on the weighted D-S evidence theory was proposed for target recognition in the distributed multi-senser system. First, the analysis of the system＇s formatlon model showed that the decision credibility of a sensor was jointly determined by two dements, namely the degree of support and the distance between it and the target. As the decision credibility can be presented by the corresponding weight of the weighted D-S evidence theory, it was obtained by a formula based on the two dements. FinaLly, the encouraging experimented results demonstrate the effectiveness of the method.
出处
《计算机应用》
CSCD
北大核心
2007年第1期56-57,97,共3页
journal of Computer Applications
