Network condition based adaptive control and its application to power balancing in electrical grids

To maintain a reliable and stable power grid there must be balance between consumption and production. To achieve power balance in a system with high penetration of distributed renewable resources and flexible assets, these individual system can be coordinated through a control unit to become part of the power balancing effort. Such control strategies require communication networks for exchange of control loop information. In this work, we show how a congested communication network can have a dramatic impact on the control performance of such a power balancing controller. To alleviate potential stability issues and increase control performance, an adaptive control design is proposed together with a communication network state estimation algorithm. Extensive simulation studies on a realistic model of a low voltage residential grid, using network traces obtained from a real powerline network, show significant performance improvement when the adaptive controller is used.