Combat Intention Recognition of Aerial Battlefield Targets Based on LSTM

In the course of military operations, the battlefield situation changes rapidly, and the combat time is fleeting. How to quickly and accurately identify the enemy’s combat intention is one of the important preconditions to ensure the commander’s correct command decision and obtain the initiative in the battlefield. In this paper, aiming at the problems of low accuracy and insufficient algorithm robustness of traditional target combat intention recognition methods in air battlefield, a combat intention recognition method based on LSTM is proposed according to the characteristics of sequence and multi-dimension of target characteristic data. Based on the historical and current state characteristics, electromagnetic characteristics and action characteristics of the target obtained by the sensor, the method uses the LSTM network to build the target combat intention recognition model, and learns the changes of the target characteristics through the network model to realize the target combat intention recognition. Simulation results show that the proposed method can effectively improve the accuracy of target combat intention recognition.