A prediction model for the management of grids operating with renewable energy sources



Researchers in South Korea have developed a forecasting model to better manage power grids with a large amount of renewable energy sources. The model was tested using historical data from the Pennsylvania-New Jersey-Maryland (PJM) grid in the US and was shown to accurately predict the availability of renewable energy resources up to a day in advance.

Energy storage is a big part of the solution to this, but other approaches to managing the production of renewable resources can also be valuable, and could be even more effective with the added ability to predict the output of a renewable energy system even just a few hours in advance.

And this is what researchers led by South Korea’s Chung-Ang University set out to achieve by developing a forecasting model for predicting renewable energy production and other uncertain parameters that help to use the grid optimally and profitably for energy producers.

“The proposed data-driven prediction method uses a long-term short-term memory (LSTM) model, an artificial neural network with feedback connections. Its hyperparameters are optimized by the genetic algorithm adaptive weighted particle swarm (GA-AWPSO) algorithm, while the global attention mechanism (GAM) identifies important features from the input parameter data,” explains Chung-Angin Professor Mun Kyeom Kim describing mathematics. behind his model. “Both algorithms can help overcome the limitations of traditional methods and improve the prediction accuracy and efficiency of the LSTM model.”

The model is also able to take into account the uncertainty factors on the demand and price side. Its operation is fully described in the article A novel deep learning-based prediction model optimized with a heuristic algorithm for microgrid energy management, published in Applied energy.

The researchers tested their work using data from part of the PJM Interconnection Network, which spans the United States, and found that it was more accurate than existing network models, particularly in predicting solar energy production. “It accelerates the integration of renewable resources into power grids while allowing MG operators to solve next-day energy management issues,” Kim said. “Ultimately, it could open the door to zero-emission electricity sources, which could make carbon neutrality by 2050 a realistic goal.”

David is a passionate writer and researcher who specializes in solar energy. He has a strong background in engineering and environmental science, which gives him a deep understanding of the science behind solar power and its benefits. David writes about the latest developments in solar technology and provides practical advice for homeowners and businesses who are interested in switching to solar.

Read More

Related Articles


Please enter your comment!
Please enter your name here