Time-Varying Nonholonomic UAV Formation Control with Trajectory Prediction and Nonlinear Model Predictive Control

Abstract

This paper proposes a time-varying unmanned aerial vehicle (UAV) formation control method based on trajectory prediction and nonlinear model predictive control (NMPC). First, in formation control, the nominal controller is constructed using a linear model and consensus for theoretical stability assurance. The time-varying formation control of nonholonomic UAVs subsequently is achieved by integrating the nonholonomic model with the NMPC, using the desired state derived from the nominal controller. Furthermore, considering the optimal obstacle avoidance problem of moving obstacles, the transformer is used to predict the trajectory in the predictive horizon, the safety constraints are established in combination with the discrete control barrier function (DCBF), and the optimal obstacle avoidance is realized by reducing the additional motion generated during obstacle avoidance. Subsequently, the NMPC-DCBF-transformer is integrated to realize the optimal obstacle avoidance control of nonholonomic UAV time-varying formation. Finally, the algorithm’s effectiveness was verified by numerical simulation, and the advantages of the algorithm were verified by comparison.