Distributed receding horizon control of cost coupled systems

This paper considers the problem of distributed control of dynamically decoupled systems that are subject to decoupled constraints, while their states are coupled non- separably in the cost function of an optimal control problem. A distributed receding horizon control (RHC) algorithm is presented, in which each subsystem (agent) computes its own control locally. The implementation presumes synchronous parallel updates. Coupled agents, referred to as neighbors, coordinate by the exchange of an assumed state trajectory prior to each update. The algorithm is an improvement over the previous algorithm developed by the author, in that stability is guaranteed without adding any additional constraints to each local optimal control problem. Instead, a move suppression term is added into each local cost function, which penalizes the deviation of the computed state trajectory from the assumed state trajectory. Closed-loop stability follows if the weight on the move suppression term is larger than a parameter that bounds the amount of coupling in the cost function between neighbors.