Cooperative Tracking of Quadrotor UAVs Using Parallel Optimal Learning Control

The cooperative tracking control issue of quadrotor unmanned aerial vehicles (UAVs) is investigated, where a cluster of UAVs is required to maintain a preassigned pattern while tracking a reference trajectory. Since only a part of UAVs can access the reference trajectory, an adaptive distributed estimator is developed such that each UAV could obtain the accurate estimate exponentially. Based on the hierarchical development, a parallel optimal learning control strategy is proposed by introducing virtual artificial systems that generate the practical control commands. In particular, a saturated force command is exploited for the position loop tracking to the estimated trajectory and a torque command is utilized to ensure the command attitude tracking, respectively. Moreover, a data-based learning law is designed for the critic weight under the finite excitation (FE), which is made available for the inadmissible initial control. It is shown that the overall closed-loop system is uniformly ultimately bounded and the tracking errors eventually converge to small sets around zero. Simulation and experiment results further verify the proposed control strategy. Note to Practitioners-This paper is inspired by the cooperative formation control issue of quadrotor UAVs. In practical missions like surveillance and reconnaissance, a UAV comes into or leaves out of the cluster may lead to re-tune the control parameters for all the involved UAVs. To save the onboard resource, the communication topology among the UAVs cannot be tedious. The proposed parallel learning strategy is implemented as follows. First, we design an adaptive distributed estimator for each UAV that could obtain the reference information accurately. Then, we propose a data-based parallel optimal learning strategy for each UAV that tracks the estimated reference trajectory. Since the proposed strategy does not require any global information regarding the communication topology, the parameters can be remained when the UAV cluster needs to extend or reduce, which makes it more practical. Finally, the proposed strategy is validated by both simulation and experiment results.