摘要
针对D-S证据理论在多传感器目标识别系统中对证据高度冲突时会产生有悖于常理的结果的问题进行了研究,对现有的方法基础上给出了一种改进的融合算法。该算法通过引入证据可信度门阈值提高了证据合成的可靠性和合理性,可得到更为理想的合成结果。最后通过算例比较表明:该改进算法的有效性。
A problem which can lead to conflict between the result that gets from D-S evidence theory when the evidence is highly conflicted and the logical result in a multiple sensor target recognition system is researched. A modified fusion algorithm is given based on the present method. The new algorithm improves the reliability and rationality of the evidence synthesis by introducing evidence credibility threshold value and more ideal synthesis result is got. By contrasting the examples ,it shows the availability of the modified algorithm.
出处
《传感器与微系统》
CSCD
北大核心
2012年第7期14-16,共3页
Transducer and Microsystem Technologies
基金
广西教育厅科研立项项目(200911LX98)
关键词
证据理论
权重系数
证据距离
组合规则
evidence theory
weight coefficient
evidence distance
combination rules
