Inferring connectivity model from meter measurements in distribution networks

We present a novel analytics approach to infer the underlying interconnection between various metered entities in a radial distribution network. Our approach uses a time series of power measurements collected from different meters in the distribution grid and infers the underlying network between these meters. The collected measurements are used to set up a system of linear equations based upon the principle of conservation of energy. The equations are analyzed to estimate a tree network that optimally fits the time series of meter measurements. We study experimentally the number of measurements needed to infer the true underlying connectivity with the help of both synthetic and real smart meter measurements in the noiseless setting.