应用D-S证据理论的雷达工作模式特征层融合识别  

A Feature-Level Fusion Recognition Algorithm of Radar Operation Mode Based on D-S Evidence Theory

作　　者：王星[1] 王志鹏[1] 呙鹏程 周东青[1] 杜文红[1] 王超[2] 

机构地区：[1]空军工程大学航空航天工程学院 [2]95959部队

出　　处：《空军工程大学学报（自然科学版）》2016年第4期35-40,共6页Journal of Air Force Engineering University(Natural Science Edition)

基　　金：航空科学基金(20152096019;20145596025)

摘　　要：日趋复杂的战场电磁环境对雷达工作模式识别提出了更高的要求,针对传统单平台工作模式识别方法的局限性,提出了一种在多平台协同背景下基于Dempster-Shafer(D-S)证据理论的雷达工作模式特征层融合识别算法。该算法运用隶属度函数获得基本信任赋值,根据Dempster合成规则,首先对多平台侦收的同一参数数据进行多平台参数内融合,然后进行参数间融合获得合成信任度,最后依据工作模式判定规则识别雷达工作模式。通过仿真实验,多平台融合识别算法的识别率较单平台识别方法平均提高了27%,实验结果说明融合识别方法有效地提高了识别性能。Aimed at the problems that the electromagnetic environments are becoming more complex day by day,and the radar operation mode recognition is of the limitations,a feature-level fusion recognition algo-rithm of radar operation mode based on D-S evidence theory is presented to overcome the limitations of rec-ognition method on single-platform under condition of the background of multi-platforms cooperative com-bat.Firstly,the BPAF （Basic Probability Assignment Function）is gotten with the member function. Then,according to the Dempster fusion rule,the fusion among platforms and the fusion among parame-ters are carried out successively.Finally,the radar operation mode is recognized by fusion results and deci-sion rule of radar mode.Through the simulation experiments,the recognition rate of multi-platform is av-eragely increased by 27%.The result shows that the method can improve the performance of recognition effectively.

关 键 词：雷达工作模式识别 信息融合 D-S证据理论 特征层 

分 类 号：TN911.7[电子电信—通信与信息系统]