Knowledge-Driven Target Trajectory Prediction Method for BVR Air Game Mission

Aiming at the high-precision real-time prediction requirements of target trajectory prediction in beyond-visual-range (BVR) air game mission, this paper proposes a method driven by aeronautical domain knowledge of BVR game missions. The target historical intentions and maneuvers are identified based on 2 different time windows respectively. Considering that the target's intention in the BVR game missions does not change drastically in a short time, the future maneuvers are predicted based on the historical intentions and maneuver identification results, and then the predicted future maneuver information is used to assist in realizing target trajectory prediction by long short-term memory (LSTM) neural network. The proposed method was tested based on BVR game mission data collected in a highly realistic simulation environment. The results showed that the method has high prediction accuracy and real-time performance.