Time-critical coordination control of multiple unmanned aerial vehicle with absolute temporal constraints

Abstract

This thesis addresses the problem of time-critical coordination of multiple Unmanned Aerial Vehicles (UAVs) with absolute temporal constraints. In this regard, the work presented here proposes a definition for strict and relaxed absolute temporal constraints, and illustrates their practicality in several aerospace missions, namely, a sequential auto-landing scenario and the calibration and validation of an Earth-observing satellite instrument. In addition, we propose a leader-follower time-critical coordination law that can support the execution of multi-vehicle missions that impose both strict absolute and relative temporal specifications on the trajectories of the vehicles. This thesis also derives stability and convergence guarantees for the closedloop coordination dynamics with the algorithm for strict absolute temporal constraints. In particular, we derive a guaranteed rate of convergence of the collective dynamics as a function of the Quality of Service of the communication network, the size of the fleet, and the number of leader vehicles. The expression obtained is consistent with previous results in time-coordination with relative temporal constraints. This work also contains a second distributed coordination control law that enforces relaxed absolute temporal assignments on the trajectories of the vehicles, in addition to the relative temporal specifications. The algorithm for relaxed absolute temporal constraints is the result of a modification of the coordination control law for strict absolute temporal specifications, where a set of state-dependent link weights are key to guarantee that the fleet reaches its final destination within a pre-defined arrival window. The proposed cooperative strategies yield robust behavior against external disturbances by allowing the vehicles to negotiate their speeds along the paths in response to information exchanged over a supporting inter-vehicle communication network. Additionally, the distributed coordination control algorithms proposed here extend the range of possible time-critical coordination missions for autonomous vehicles addressed in previous work. Simulation results of a sequential auto-landing scenario with five UAVs demonstrate the efficacy of the control laws for group coordination presented in this thesis.