Safe Adaptive Dynamic Programming-Based Formation Control and Collision Avoidance of Multi-UAV System

In this study, a safe adaptive dynamic programming-based method is proposed for the formation control and collision avoidance problem of the multi-UAV systems with the known upper bound of external disturbances. For the position subsystem, an adaptive dynamic programming algorithm is established, using a designed cost function that manages to include uncertainty, tracking error, control influence, and barrier functions. Through the critic network structure with a unique weight learning criterion, the optimal cost function and control strategies are obtained. After attitude resolution, an integral back-stepping controller is established for the attitude subsystem. The simulation experiment is designed to verify the effectiveness of the algorithm.