基于证据理论的战场被动声多目标识别研究  

Research on Recognition of Passive Acoustic Multitarget Based on Evidence Theory

作　　者：陈丹[1] 李京华[2] 黄根全[3] 许家栋[2] 

机构地区：[1]西安理工大学自动化与信息工程学院,陕西西安710048 [2]西北工业大学电子信息学院,陕西西安710072 [3]西安电子工程研究所,陕西西安710100

出　　处：《系统仿真学报》2007年第6期1323-1325,共3页Journal of System Simulation

基　　金：国防重点实验室基金(51454070204HK0320)

摘　　要：Dempster组合规则是证据推理的重要部分,广泛应用在多传感器信息融合算法中。组合法本身存在一些不足之处,应用到战场多目标识别中不能达到满意的识别率。分析了两种改进组合方法的基本理论和特点,应用到战场多类声目标识别的特征融合实验中均获得满意的识别结果。验证了两种改进组合方法在战场目标识别中具有一定的应用前景。Dempster＇s rule is an important part of evidence reasoning, which has been widely used in multi-sensor data fusion algorithms. But Dempster＇s rule used in recognition of Passive acoustic multi-target could not get satisfactory recognition rate for it＇s own drawbacks. The concepts and characters of two improved combination rules were introduced. The high recognition rates were got after they were applied to acoustic multi-target recognition. The experimental results show two improved combination rules in the filed of recognition of battlefield multi-target have certain applied prospect.

关 键 词：证据推理 组合规则 目标识别 识别率 

分 类 号：TN911[电子电信—通信与信息系统]