Structure Learning in Power Distribution Networks

摘要: Traditional power distribution networks suffer from a lack of real-time observability. This complicates development and implementation new smart-grid technologies, such as those related to demand response, outage detection management, improved load monitoring. In this paper, inspired by proliferation metering technology, we discuss topology estimation problems in structurally loopy but operationally radial grids measurements, for example, voltage data, which are either already available or can be made with relatively minor investment. The primary objective paper is learn the operational layout grid. Further, structure learning algorithm extended cases missing where observations limited fraction grid nodes. algorithms computationally efficient—polynomial time—which proven theoretically illustrated numerical experiments on number test cases. techniques developed applied detect line failures real time well understand scope possible adversarial attacks