基于深度学习的来袭鱼雷态势判断方法

Research on torpedo trajectory for vessels based on deep learning

鱼雷防御是舰艇安全执行任务的重要前提。准确判断来袭鱼雷的态势是成功防御的关键。错误的态势判断和对抗方法会降低鱼雷防御效果,甚至增加被命中的可能性。因此非常有必要研究来袭鱼雷的态势判断方法,从而为准确制定防御方法提供依据。由于不同类型的鱼雷有不同的运动模型和弹道形态,为此,研究和建模来袭鱼雷的运动模型是一种准确判断鱼雷类型的有效途径。本文在综合调研的基础上,分别对常见的几种舰艇鱼雷运动模型进行分析、建模和仿真,并基于长短期记忆人工神经网络对仿真结果进行了分析和验证。该研究可为舰艇鱼雷防御决策提供重要参考。

Torpedo defense is a critical prerequisite for vessels to perform their missions safely.The accurate judgment of the torpedo type is one of the important factors for the success of countering incoming torpedoes.Incorrect judgment may guide commander to adopt a wrong countermeasure,which will greatly reduce the effect of torpedo defense.More seriously,the wrong action will increase the risk of exposure of the vessels and further be hit.Therefore,it is very necessary to study the types of incoming torpedoes to provide a basis for accurately formulating methods to counter torpedoes.Because different types of torpedoes have different types of trajectories,investigation and modeling the trajectory of incoming torpedoes is an effective way to determine the type of torpedoes.This paper analyzes models and simulates several common vessels torpedo trajectories,which can provide an important reference for vessels torpedo defense decisions.