This paper proposes a decentralized conflict detection and resolution method using probabilistic trajectory prediction combined with intent inference. In our method, by incorporating the results of inverse optimal control for intent inference and the unscented transform for predicting nonlinear probabilistic state propagation, conflict detection and resolution problem is formulated as a decentralized model predictive control (MPC) problem. The method can cover linearonlinear uncertain behaviors of aircraft due to the wind effect, the deviation of navigation and control strategy, and the inaccuracy of inferred intent. In addition, it can circumvent nonconvex constraints representing the separation condition as well as flight performance limits in computationally tractable way by decoupling and convexification technique. Through numerical simulations, the effectiveness of the method is demonstrated in terms of the validity of the obtained conflict-free trajectories and the computational efficiency.
展开▼