An Intention Recognition Method Based on Bayesian Network for UAV Formation in Confrontation

It's important to predict the intention of the UAV formation for both side of aerial confrontation to gain operational advantages, for the combat of UAV formation confrontation is dynamic and changeable. Aiming at the problem of the intention recognition for the UAV formation in combat, this paper proposed a method based on dynamic Bayesian networks. First, the UAV formation description model is used to describe the speed, altitude and geometry configuration of the leader-follower UAV formation. Then, the effect of combat result was analyzed by configuration and state attributes of the UAV formation during aerial confrontation. Finally, the combat intentions were predicted based on dynamic Bayesian networks with formation flight data. The simulation results showed that the dynamic Bayesian-based approach could effectively identify the adversarial intention of UAV formation.