Weather-Dependent Fast Power Flow in Distribution Systems: A Meteorological-Data-Aided Method

The weather-dependent variability of line impedance can substantially impair the accuracy of traditional power flow calculations in electric grids, especially under varying and extreme weather conditions. Recently, significant advancements have been made in weather-dependent power flow (WDPF) algorithms to address this issue in transmission systems. However, determining line parameters in three-phase unbalanced distribution grids remains time-consuming due to the nonlinear and analytically intractable nature of heat-balance functions. To achieve fast and high-fidelity calculations, this paper proposes a novel data-aided WDPF algorithm leveraging readily available weather information to calibrate line parameters in real time. The method enhances the classic backward-forward sweep approach for unbalanced distribution grids by incorporating real-time updates of line parameters based on the thermal model of line conductors during the iterative power flow process. Additionally, well-trained neural networks are employed to expedite power flow calculations by solving line temperature variables. Case studies on two unbalanced IEEE feeders demonstrate that the proposed algorithm surpasses state-of-the-art WDPF algorithms in computational efficiency and accuracy under diverse weather conditions.