基于粗糙自适应神经模糊推理的导弹类型识别  

Missile Type Recognition Based on Rough Adaptive Neural-fuzzy Inference Method

作　　者：李大鹏[1] 杨根源 李京[1] 

机构地区：[1]海军航空工程学院,山东烟台264001 [2]海军信息化专家咨询委员会,北京100073

出　　处：《战术导弹技术》2012年第4期30-34,共5页Tactical Missile Technology

摘　　要：舰艇编队的反导作战能力直接关系着编队的生存能力,而导弹目标类型识别是反导作战的基础。针对舰艇编队反导作战对导弹类型识别的具体战术要求,建立了反舰导弹分类及特征模型,分析了编队反导作战目标类型综合识别流程。结合粗糙集理论与自适应神经模糊推理,设计了一种基于粗糙自适应神经模糊推理(RAN-FIS)的导弹类型识别方法,提高了识别效率。根据LMS(least-mean square)算法设计了RANFIS的参数估计方法,该参数估计方法根据实际输出信号与网络输出信号之间的误差调整参数取值,能够根据已有反舰导弹样本的特征属性以及类型确定参数。仿真结果显示,所设计的导弹类型识别方法能够很好的识别海战场上出现的反舰导弹的类型。The anti-missile fight capability concerns the fleet vitality, and missile type recognition is the base of anti-missile fight. In view of the tactics require of the warship formation anti-missile fight missile type recognition, the classification and characteristic model of anti-ship missile were constructed, the flow of missile type recognition was analyzed. A recognition method based on rough adaptive neural-fuzzy in- ference system was designed, which improve the recognition efficiency. The missile type recognition process was analyzed, the parameter estimation method was designed based on LMS （least-mean square） algorithm, which can adjust parameter based on the error between actual output signal and the designed network output signal. Simulation show that the designed missile type recognition method can ascertain the type of missiles which appear on the sea battlefield.

关 键 词：舰艇编队 反导作战 导弹类型识别 粗糙自适应神经模糊推理系统 

分 类 号：TP182[自动化与计算机技术—控制理论与控制工程]