Model predictive control for conventional power production optimization and voltage rise mitigation in distribution networks

Abstract

The energy flow management in distribution networks started to worry grid operators since the challenge of exploiting natural resources such as solar irradiance and wind speed to produce electricity became a reality. Due to their intermittent nature, mismanagement of these resources can result in a severe bidirectional energy flow; likewise, notable imbalances between demand and generation in the electrical lines. Consequently, this may engender excessive energy losses and could violate electrical component characteristics. However, a predictive management of these resources can allow for the forecast of the different existing electrical device regulations beside the energy needed to be produced by conventional generations in a vertically-integrated system. For that purpose, we present in this paper a Model Predictive Control approach applied on a distribution feeder connected to renewable distributed generators and supplied by a High Voltage/Medium Voltage substation through transmission lines. We associated to this model a regulating algorithm for energy optimization and cost minimization. The control strategy outputs are the active and reactive energy predicted to be injected by the substation into the feeder. Finally, we evaluated the accurateness of the proposed method and we identified the parameters that affect its precision error.