Lagrange Multipliers based price driven coordination with constraints consideration for multisource power generation systems

In the context of distributed Model Predictive Control (MPC), which has been attracting more and more attention in the recent years, the present paper proposes a price-driven coordination approach that allows explicit and decentralized computation of a model predictive controller for large scale systems (in the spirit of [1]), but also handling constraints. Based indeed in the relaxation and separation of a performance cost function, the coordinator computes a price vector (Lagrange Multiplier) that coordinates the global system while respecting its constraints. This vector is transmitted to the subsystems that use this information for adjusting the control policies according to the global objectives, but without total dependence on the coordinator. In particular, if this information is lost, the system can still be operative because the subsystems control strategies are computed locally. In this frame, a distributed power system is considered as an example, in which it is shown that the proposed technique indeed maintains the system in a stable operation, with the same characteristics as the case of a centralized control solution, as well as robustness under loss of information from the coordinator.