基于Transformer-LSTM网络的干扰态势预测

打开文本图片集

中图分类号：TP183；TN957.51 文献标志码：A

Abstract： Focusing on two types of combat units：radar and jammers，the Transformer-Long Short Term Memory（Transformer-LSTM） network is employed to predict the operational parameters and modes of radars，as wellas the jamming patterns of jammers.By combining the advantages of LSTM and Transformer models，the method enhances the ability to learn complex patterns in sequential data.Reinforcement learning is used to simulate the generation of enemy jamming patern time-series datasets based on our radar operational status dataset，making the prediction results more closely aligned with actual combat scenarios.By introducing deep learning and reinforcement learning techniques，the prediction capability for the operational status of radars and jammers is improved，achieving significant tasks in situational prediction，such as forecasting the operational states of radars and jammers.

KeyWords：situation prediction；reinforcement learning;LSTMnetwork；Transformer model

0 引言

雷达干扰态势预测是态势认知系统中的重要组成部分。（剩余10320字）