Big Data for Accurate Predictions of Electricity Grid Flow
Big data software solutions to improve the accuracy of short-term predictions of the flow of electricity through grids.
Professor José Vilar, Spain
One of the challenges of delivering electricity through the grid nowadays is that it is no longer produced by a handful of stable plants. Instead, it comes from millions of dispersed and unpredictable sites such as wind turbines and solar cells. This makes it more difficult to ensure that electricity supply matches demand at all times. Fortunately, there is a positive way forward, and that is with big data.
Leveraging Electricity Big Data
Energy grids are complex and interconnected and consequently generate colossal amounts of data such as by power lines, at turbines and places where the energy is consumed. Leveraging that data can help to make accurate short-term predictions about power grid traffic. This knowledge can then help to reduce the risk of blackouts or energy wasted. But what is the best way to turn that data into something useful?
That was the question being asked of participants of the Big Data Technologies Horizon Prize in 2018. The open innovation contest was open to groups, individuals and organizations taking part in the European Union's research and innovation program, Horizon 2020.
They were tasked with using big data to develop the most accurate software to predict the likely electricity flow through a grid taking into account several factors. These included the weather and the generation source. Data from the grids was combined with additional data such as the weather conditions.
Carlos Moedas, Commissioner for Research, Science and Innovation, said: “Energy is one of the crucial sectors that are being transformed by digital technologies. This Prize is a good example of how we support a positive transformation through the EU's research and innovation program, Horizon 2020. For the future, we have designed our next program, Horizon Europe, to put even more emphasis on the merger of the physical and digital worlds across sectors such as energy, transport and health.”
The software models had to accurately predict the flow of energy through the grid over a six-hour period, and they were tested for speed and accuracy. Of the 24 applications, seven managed the task successfully. From these seven, three individuals were selected as winners based on the combined rank of accuracy and speed with greater weight being given to accuracy.
The first prize of €1.2 million (approximately $USD 1.3 million) went to Professor José Vilar from Spain, and €400,000 to Belgians Sofie Verrewaere and Yann-Aël Le Borgne who finished in joint second place.
Organizers of the open innovation big data contest hope that the energy sector will take up the winning solutions and further develop them to ensure more economic and more reliable power grids.
"The wide range of possible applications of these winning submissions could bring tangible benefits to all European citizens,” commented Mariya Gabriel, the EU Commissioner for Digital Economy and Society.
Next Story »