Predicting Renewable Curtailment in Distribution Grids Using Neural Networks

The growing integration of renewable energies into electricity grids leads to an increase grid congestions. One countermeasure is the curtailment energies, which has disadvantage wasting energy. Forecasting congestion provides valuable information for operators prepare and instruct countermeasures reduce these energy losses. This paper presents a novel approach prediction in distribution (i.e. up 110 kV) considering n-1 security criterion. For this, our method considers node injections power flow combines three artificial neural network models. analysis study results shows that implemented networks within presented perform better than naive forecasts In case vertical flow, also show comparable parametric models: average values mean absolute errors relative models range from 0.89 0.21. A high level accuracy can be achieved predicts loading components with F1 score 0.92. Further, 0.92, this model higher those transmission grid, achieve 0.84. approaches good potential support management.