TLSTM: A Transformer-LSTM Method for UAV Combat Intent Recognition

Drones have become an indispensable tool in modern information-based combat systems. In the Russian-Ukrainian conflict, both sides have extensively utilized unmanned aerial vehicles (UAVs) for high-intensity offensive and defensive confrontations, highlighting the crucial role of drones in contemporary warfare. This paper focuses on exploring the practical applications of UAV technology and delves into the intricacies of UAV intention recognition. It proposes a novel method, Transformer-Long Short Term Memory (TLSTM), to enhance the accuracy of UAV combat intention recognition. Through rigorous evaluation and comparison with various deep learning techniques, TLSTM demonstrates exceptional performance, achieving an impressive intention recognition accuracy of 98.59%. These findings have significant implications for assisting commanders in making critical decisions, thereby enhancing the effectiveness of military operations.