Data-Driven Optimal Formation Control for Multiple Nonlinear Quadrotors With Switching Topologies

This paper addresses the optimal distributed formation control of multiple quadrotor vehicles under switching topologies, where each vehicle suffers from significant nonlinearity and uncertainty, simultaneously. A distributed observer is constructed to generate the trajectory references for each vehicle. An optimal leader-following formation control protocol is then developed through the reinforcement learning method from the sampled vehicle data. Simulation results from a multiple quadrotors example under switching topologies are presented to verify the obtained results.