Sparse Overcomplete Representations for Efficient Identification of Power Line Outages

摘要: Fast and accurate unveiling of power-line outages is paramount importance not only for preventing faults that may lead to blackouts, but also routine monitoring control tasks the smart grid, including state estimation optimal power flow. Existing approaches are either challenged by combinatorial complexity issues involved thus limited identifying single double line-outages or they invoke less pragmatic assumptions such as conditionally independent phasor angle measurements available across grid. Using a subset voltage data, present paper develops near real-time algorithm multiple line at affordable solving sparse signal reconstruction problem via greedy steps coordinate descent iterations. Recognizing number small fraction total lines, novel approach relies on reformulating DC linear flow model overcomplete expansion leveraging contemporary advances in compressive sampling variable selection. This representation can be extended incorporate information internal system more general line-parameter faults. Analysis simulated tests 118-, 300-, 2383-bus systems confirm effectiveness outages.