摘要
针对航空反潜活动中声纳浮标搜潜的典型任务场景,建立了浮标布阵的搜潜仿真模型及计算搜潜成功概率的数学模型。设计了一种计算满足目标搜潜概率需用最少浮标数量的优化方法,使用遗传算法求解了该优化问题,并生成大量浮标布阵样本。构造了浮标布阵辅助决策的神经网络模块,对最优浮标布阵方式进行了训练与学习。仿真验证表明,使用该方法能有效满足搜潜辅助决策任务需求。
For typical cases in aviation anti-submarine task, both simulation model of sonobuoy array in searching submarine and mathematic model of calculating success probability are established. An optimal solution method of using minimum sonobuoys is designed, which satisfies searching task objectives and is solved by genetic algorithm. A large amount of samples are generated. Furthermore, by introducing artificial intelligence method, the optimal sonobuoy arrangement samples are trained and learned, thus a neural network module for sonobuoy array assistant decision-making is built. Simulation examples show this method meets the needs of decision-making tasks in searching submarine.
引文
[1]邓歌明,周晓光,冯百胜,等.基于对策论的舰载反潜机反潜作战兵力部署研究[J].火力与指挥控制, 2017, 42(4):48-50.
[2]韩海良,韩瑜.基于BP神经网络的水面舰艇CGF反潜指挥决策模型[J].指挥控制与仿真, 2014,36(1):63-66.
[3]李居伟,宋旭,孙明太,等.反潜巡逻机使用浮标声纳和自导深弹攻潜模型仿真[J].弹箭与制导学报,2011,31(1):219-222.
[4]蒋志忠,杨日杰,李大龙,等.反潜巡逻机布放包围浮标针应召搜潜建模与仿真[J].海军航空工程学院院报, 2011, 26(4):422-426.