A distributed economic MPC framework for cooperative control under conflicting objectives

In this paper, we consider the problem of coordinating self-interested interacting dynamical systems by means of a distributed economic MPC framework. The self-interest of the systems is reflected by an individual local objective function each agent is trying to minimize, while cooperation is required with respect to coupling constraints and an asymptotic cooperative goal, which is represented by a particular steady state of the overall system. Our basic premise is that this steady state, which fulfills the cooperative goal, is not known a priori but has to be negotiated online, while already taking control actions. For the purpose of determining this steady state in a distributed way, arbitrary distributed computation algorithms can be incorporated into the proposed framework. We show that satisfaction of coupling constraints and convergence to the desired overall steady state can be established. Examples for an asymptotic cooperative goal include synchronization under conflicting objectives or sensor coverage, which are both studied in the work at hand and are also illustrated by numerical simulations.