Find in Library

In information-based air combat, whoever can quickly obtain and effectively utilize opponent information, can accurately predict the opponent’s movement trajectory, complete the OODA cycle more quickly, and thus gain air combat advantages. This article studies the trajectory prediction technology of air combat targets from the characteristic of time series variation, analyzes and summarizes trajectory prediction technologies based on physics, machine learning, and deep learning. Furthermore, physical factors, air combat factors, and interaction factors are used as inputs to the trajectory prediction model. Single modal trajectory, multimodal trajectory, and behavioral intention are used as outputs of the model. Finally, this article looks forward to the target trajectory prediction technology in future intelligent air combat.