Line Parameter and Switching State Identification Method for Radial and Meshed Distribution Networks

Current major changes in distribution systems require better monitoring from energy utilities to prevent possible issues in the form of deterioration of power quality and stability, voltage oscillations, and many others. However, distribution networks are still characterized by their limited observability, where information such as switching state and line parameters are not known. This constraint further complicates analyses concerning the observations of distribution networks. Therefore, this paper proposes a methodology that, as an answer to these problems, offers machine learning-based algorithms that aim to determine the switching state and line parameters of distribution networks using only collected historical Advanced Metering Infrastructure (AMI) measurements. This model was tested on the radial IEEE 33-bus benchmark system and both meshed and radial real-world distribution networks. The results demonstrate that the proposed method can provide an accurate estimation of line parameters and network topology relying only on voltage magnitudes and active and reactive power measurements.