Webinar

Abstract:

Distributed networks, like power systems, call for distributed, but coordinated, control techniques. These allow to combine the advantages of one centralized controller and fully-autonomous controllers for each component, while avoiding some of their respective disadvantages. One approach to achieve this is by solving the optimization problem, that arises from the concept of Model Predictive Control (MPC), in a distributed way. For that purpose, a variant of the Alternating Direction Method of Multipliers (ADMM) is presented, along with a description of the arising MPC problem. The resulting control algorithms are then empirically evaluated in a case study, where they are applied to different models of power networks in different scenarios. Additionally, the controller is combined with an Extended Kalman Filter, that estimates the inner system state from a reduced set of measurements.