Deep Learning loss model for large-scale low voltage smart grids

摘要: Abstract Distribution systems operators (DSOs) encounter the challenge of managing network losses in large geographical areas with hundreds secondary substations and thousands customers an ever-increasing presence renewable energy sources. This situation complicates estimation process power loss, which is paramount to improve efficiency level context European Union policies. Thus, this article presents a methodology estimate large-scale low voltage (LV) smart grids. The based on deep-learning loss model infer technical considering rollout meters, high penetration distributed generation (DG) unbalanced operation, among other characteristics. has been validated LV distribution area Madrid (Spain). proposed proven be potential tool grids resources. accuracy outperforms that state-of-the-art methods, exhibiting rapid convergence allows for its use real-time operations.