The use of constraint programming for the autonomous management of power flows

This paper describes a novel network agnostic approach to the real-time management of power flows within MV distribution networks with a number of embedded generators. The approach in question uses the artificial intelligence technique of constraint programming to autonomously make decisions about how to manage a number of generators as part of a larger autonomous regional active network management system called AuRA-NMS. The results of testing using commercially available IEC 61850 compliant substation computing hardware, connected to a realtime simulator, are presented for two different case study networks: an interconnected 33 kV network and an 11 kV radial network. These results demonstrate the feasibility of the approach both in terms of its applicability to different network topologies and its real-time performance. The paper concludes by discussing the methodology for future testing which aims to assess the robustness of the approach against a number of other proposed approaches and the plans for field trials of the current prototype.