摘要
针对平台型号融合识别问题,依据电子支援措施(ESM)和雷达情报源,利用载频、脉冲间隔和脉宽等信号参数特征,以及速度、高度和雷达散射截面(RCS)等状态参数特征,再利用活动区域、机场位置和飞行模式等先验信息有关的目标位置和航迹特征,提出了能获得信号级、状态级和态势级分类信息的3种单分类器。在此基础上,设计了基于加权和规则的融合分类器。仿真验证结果表明,融合分类器的分类性能优于各单分类器,能充分发挥ESM和雷达情报多种参数以及先验配置信息在识别中的作用,整体上提高了目标识别性能。
To solve a platform type fusion recognition problem, three kinds of single classifiers for obtaining sorted information of signal level, state level and situation level are proposed based on the electronic support measure （ESM） and the radar intelligence source. Signal parameter characteristics, such as carrier frequency, pulse interval and pulse width~ state parameter charac- teristics, such as speed, height and radar cross section （RCS）, and situation parameter character- istics, such as target position and track characteristics related to the prior information of activity area, airport position, and flight patterns are used in three single classifiers. The fusion classifier is designed based on weighted sum rules. Simulation results show that, the performance of the fusion classifier is better than that of each single classifier. The target recognition performance is improved by using prior information and the parameter characteristics of ESM and radar intelli- gence.
出处
《指挥信息系统与技术》
2013年第6期54-58,共5页
Command Information System and Technology
基金
国家自然科学基金(61271144)资助项目
关键词
多分类器
目标识别
电子支援措施
雷达情报
multiple classifiers
target recognition
electronic support measure （ESM）
radar in- telligence
