Artificial Neural Networks Proven to Speed Up Power Flow Calculations by up to 45x

Strategic Insights -

Researchers at Universitat Politecnica de Catalunya have successfully utilized artificial neural networks (ANNs) and XGBoost techniques to simulate power flows in a medium-voltage grid in Norway with significant speed-up factors. This groundbreaking study, supported by the European Union, demonstrates the potential of machine learning (ML) models in managing power systems and ensuring reliability and efficiency.

Key Takeaways:

  • The research utilized a case study where ANNs and XGBoost were trained to simulate power flows in a medium-voltage grid in Norway, estimating the impact of battery energy storage systems (BESSs) on grid performance.
  • The proposed methodology can be used to assess the impact of other grid-connected assets, such as small-scale solar plants and electric vehicle chargers, on distribution networks.
  • The study proved that while ML models require considerable data and training time, they offer speed-up factors of up to 45x, depending on the predicted parameter.
  • The research highlighted the potential of machine learning (ML) models in managing power systems and ensuring reliability and efficiency.
  • The study presented a simulation of power flows in a medium-voltage grid in Norway, using thousands of configurations to demonstrate the speed-up factors of up to 45x.
  • The research suggested that the proposed methodology can be used to evaluate the effects of distributed energy resources on grid performance.

Statistics:

  • The speed-up factors of up to 45x were achieved by using ANNs and XGBoost techniques in simulating power flows in a medium-voltage grid in Norway.
  • The study used a case study with thousands of configurations to demonstrate the speed-up factors of the proposed methodology.
  • The research highlighted the potential of machine learning (ML) models in managing power systems, ensuring reliability, and increasing efficiency.

Sources:

  • Energies, "Evaluation of XGBoost and ANN as Surrogates for Power Flow Predictions with Dynamic Energy Storage Scenarios," 2025,18(16):4416. (Energies - http://www.mdpi.com/journal/energies).
  • DOI: https://doi.org/10.3390/en18164416.