基于LSTM的导弹拦截点预测研究

Research on Missile Interception Point Prediction Based on LSTM Network

在导弹突防过程中,对拦截弹拦截点和拦截时间的预测能够为成功突防提供有力支持。因此,针对拦截弹拦截巡航弹的攻防场景,提出一种基于LSTM的神经网络模型对拦截时间和拦截点进行预测。通过攻防仿真获得拦截弹弹道数据作为训练集对网络进行训练,模型以拦截弹的位置作为输入量,用神经网络估计拦截弹轨迹类型,得出拦截点和拦截时间。实验结果表明,该网络能够有效预测拦截点和拦截时间,且预测误差较小,可为导弹突防策略提供有效的参考。

In the process of missile penetration,the prediction of interceptor interception point and interception time can provide strong support for successful penetration.Therefore,aiming at the attack and defense scenario of interceptor intercepting cruise missile,a neural network model based on LSTM is proposed for the prediction of interception time and interception point.The interceptor trajectory data is obtained through attack and defense simulation as the training set to train the network.The model takes the position of interceptor as the input,uses the neural network to estimate the interceptor trajectory type,and obtains the interception point and interception time.The experimental results show that the proposed network can effectively predict the interception point and interception time with small prediction error,and can provide an effective reference for missile penetration strategy.