基于矩阵对策与遗传算法的无人机空战决策  

Study on Air Combat Decision Method of UAV Based on Matrix Game and Genetic Algorithm

作　　者：邓可[1] 彭宣淇 周德云[2] DENG Ke;PENG Xuan-qi;ZHOU De-yun(PLA Dalian Naval Academy,Dalian 116018,China;Northwestern Polytechnical University,Xi’an 710072,China)

机构地区：[1]海军大连舰艇学院,辽宁大连116018 [2]西北工业大学,西安710072

出　　处：《火力与指挥控制》2019年第12期61-66,71,共7页Fire Control & Command Control

基　　金：军队“十二五”装备预先研究基金资助项目(51305080301)

摘　　要：针对无人机自主空战决策的实时性与准确性需求,提出一种矩阵对策法与遗传算法相结合的空战决策算法。基于空战优势函数以及无人机机动模型建立了无人机空战决策模型,使用改进的矩阵对策法获取我方无人机最优选择策略的大致范围,再在此范围内使用遗传算法寻找最优策略,最后通过直线飞行机动和S型规避机动仿真验证了算法的可行性。仿真结果表明,所提算法不仅能够提高机动决策的精度,还在一定程度上避免了单纯遗传算法计算量大、计算时间长的问题,满足空战合理性与实时性需求。Aiming at the real-time and accuracy requirements of UAV autonomous air combat,a decision-making air combat algorithm composed of matrix game method and genetic algorithm is put forward.UAV air combat decision-making model was built on combat advantage funtion and UAV maneuver model the improved matrix game method was adopted to get the approximate range of UAV optimal selection strategy,and then genetic algorithm was used to find the optimal maneuver selection in this range.Finally,the feasibility of the algorithm was tested by straight flight and S type evasion maneuver simulations.Simulations showed that the algorithm could not only improve the precision of maneuvering decision.But also the amount of calculation and time consuming were largely decreased compared with simply adopting genetic algorithm,which could meet the requirements of reasonability and real-time air combat.

关 键 词：无人机 自主机动决策 矩阵对策法 遗传算法 

分 类 号：V219[航空宇航科学与技术—航空宇航推进理论与工程] TJ761[兵器科学与技术—武器系统与运用工程]