Short-term load forecasting on MV/LV transformer level

Abstract

Distribution system operators (DSOs) are installing increasing amounts of measurement equipment in the distribution networks. The measurements provide the DSOs insights in the challenges distributed energy resources bring to the network. The next step is to utilise the data generated from the field measurements. This paper will introduce a practical implementation of a short-term load forecast for residential loads on medium-to-low voltage level. As this forecast is applied in a congestion management setting, the error over time is visualised, and the model is optimised for the moment at which the congestions are expected. The practical implementation of the forecasting algorithm is part of the Dutch demonstrator of the Horizon 2020 InterFlex project.