基于Rough集和回归型SVM的超视距空战威胁评估  

Threat Assessment Based on Rough Set and Support Vector Machines for Regression in Beyond-Visual-Range Air Combat

作　　者：张文忠[1] 孙永芹[1,2] 杨洪立[1] 张国贤[1] 

机构地区：[1]中国人民解放军91206部队 [2]海军潜艇学院

出　　处：《四川兵工学报》2013年第7期14-18,共5页Journal of Sichuan Ordnance

摘　　要：针对现有超视距空战威胁评估方法的不足,提出了基于Rough集和回归型SVM的超视距空战威胁评估方法。该方法结合对超视距空战过程和影响参数的分析,构造了参战双方战机的态势、效能、事件、目标战役价值等优势函数,将导弹射程作为距离因素的一种引入威胁评估优势函数模型中,改进了距离优势函数,优化了方位角、进入角、能量、效能等优势函数,并将目标战役价值作为威胁评估的因素;然后采用Rough集的知识简约方法简约了各优势函数中的冗余信息,并采用回归型SVM实现了最终的威胁评估排序;经仿真计算验证,该方法合理、有效。To cope with the shortage of current threat assessment method in beyond-visual-range air com- bat, a new threat assessment method based on rough ret and support vector machines for regression is pro- posed. The process of air combat and inflation parameters are analyzed in the method, and superiority functions such as situation, efficiency, events and the goal campaign value are constructed, aAlso, missile scope is introduced as a factor of distance. Superiority functions such as azimuth, enter angle, energy, ef- ficiency are optimized, and distance is improved. Moreover, the goal campaign value is considered as a factor of threat assessment. Then, redundancy information of each superiority function is eliminated by re- duction of knowledge based on rough set, and the final taxis of threat assessment are educed by support vector machines for regression. Finally, the rationality and effectiveness are verified by simulation results

关 键 词：超视距空战 威胁评估 优势函数 ROUGH集 知识约简 支持向量机(SVM) 

分 类 号：V271[航空宇航科学与技术—飞行器设计]