Thrust-vectoring UAVs can realize decoupling of position and attitude compared with conventional quadrotors due to the ability to change thrust direction, and are used to perform various complex indoor and outdoor missions. However, existing trajectory generation frameworks are mostly for quadrotors with fixed thrust direction and a coplanar surface, and do not consider the dynamics of thrust-vectoring UAVs. To address this, this paper proposes a multi-objective trajectory generation method for thrust-vectoring UAVs in constraint space. By parametrically modeling the constraint space, the method considers the effects of environmental boundary constraints and platform dynamics characteristics on the collision constraints and motion decoupling of the trajectory, and comprehensively optimizes the trajectory’s indicators of stability, speed, and safety to plan the states and input actions of the flight trajectory. Meanwhile, a trajectory generation evaluation system is proposed, given that compared with the conventional quadratic objective function, the proposed method is effective in reducing the attitude change of the trajectory, improving the rapidity and safety, in which L_ and Lₑ ₈ ₒ ₊ are reduced by 70. 4% and 19. 1%, respectively. Meanwhile, by comparing with the conventional quadrotor, the advantages of the thrust-vectoring in decoupling motion are quantified, especially in reducing the attitude change during flight, the pitch angle of the generated trajectory is reduced from ±30° to within ±20° degrees, which exerts the motion decoupling advantages of the thrust-vectoring.
Yongjie et al. (2026) studied this question.