Distributed Sensor-Tolerant MPC for Formation Tracking of Networked Multicopters with Input/State Constraints

This brief presents a novel model predictive control (MPC) algorithm to achieve formation tracking of networked multicopters (nodes) with sensor faults and physical constraints. The algorithm can be used to asymptotically form and maintain a user-defined three-dimensional pattern for the multicopters formation. Another concern lies on how to deal with the usual sensor faults caused by the failure of the Global Positioning System (GPS) or the Inertial Navigation System (INS), in the above-mentioned control process. To this end, we embed the residual generators and auxiliary variable observers into the algorithm for the detection and estimation of the multicopters faults. Multicopters only accept the information related to their neighbors for real-time online rolling optimization. By using the sum of local cost functions to construct a Lyapunov candidate, it is proved that the asymptotic stability of such a distributed MPC (DMPC) is guaranteed under the condition that the weight requirements are satisfied. Simulation results demonstrate the rationality and effectiveness of the proposed algorithm.