Switching function-based optimal active fault-tolerant bipartite consensus control for UAV swarm

This paper studies the problem of optimal active fault-tolerant bipartite consensus control for flying wing unmanned aerial vehicle (UAV) swarm with nonidentical and unknown direction faults (NUDFs) and disturbances by integrating switching and saturation functions. An observer–controller framework is employed to prevent error propagation among follower UAVs. Based on this framework, an active fault-tolerant control strategy is proposed to improve the system's transient performance. This strategy prevents sudden shocks from excessive reverse inputs on the system and avoids state chattering caused by excessive adjustment inputs, which often occur in the Nussbaum function-based fault-tolerant control method. Additionally, the newly proposed switching criterion directly matches the desired operating mode, even in the presence of disturbances and multiple fault direction changes, thus avoiding ineffective switching and enhancing the robustness of the designed controller. To further improve the UAVs’ transient performance, a reinforcement learning (RL)-based preset performance backstepping control method is introduced. This method optimizes output regulation, saves energy consumption, and reduces the impact of soft saturation on system performance. Finally, simulation results validate the effectiveness and superiority of the proposed scheme.