Search | Contact | Subscribe | Sign In or Register

Crisis Response Journal Crisis Response Journal

Machine learning can predict if storms will cause blackouts

Posted on 2nd August 2019 at 16:32pm

Thunderstorms are common all over the world in the summer months. Aside from spoiling afternoons in the park, lightning, rain and strong winds can damage power grids and cause electricity blackouts. It's easy to tell when a storm is coming, but electricity companies want to be able to predict which ones have the potential to damage their infrastructure. A recent collaborative study between computer scientists at Aalto University and the Finnish Meteorological Institute has applied machine learning to predict how damaging a storm might be.

Finland sees its fair share of thunderstorms, particularly in its central regions. According to new research, machine learning technology can be used to predict the severity of storms and thus help companies to foresee and mitigate for blackouts. Photo: Vasin Leenanuruksa|123rf

Machine learning - when computers find patterns in existing data which enable them to make predictions for new data - is ideal for predicting which storms might cause blackouts. Roope Tervo, a software architect at the Finnish Meteorological Institute (FMI) and PhD researcher at Aalto University in Professor Alex Jung's research group has developed a machine learning approach to predict the severity of storms.

The first step of teaching the computer how to categorise the storms was to provide it with data from power-outages. Three Finnish energy companies, Järvi-Suomen Energia, Loiste Sähkoverkko, and Imatra Seudun Sähkönsiirto, who have power grids through storm-prone central Finland, provided data about the amount of power disruptions to their network. Storms were sorted into four classes. A class zero storm didn't knock out electricity to any power transformers. A class one storm cut-off up to 10 per cent of transformers, a class two up to 50 per cent, and a class three storm cut power to over 50 per cent of the transformers.

The next step was taking the data from the storms that FMI had, and making it easy for the computer to understand. "We used a new object-based approach to preparing the data, which makes this work exciting," said Tervo. He continues: "Storms are made up of many elements that can indicate how damaging they can be: surface area, wind speed, temperature and pressure, to name a few. By grouping 16 different features of each storm, we were able to train the computer to recognise when storms will be damaging."

The results were promising: the algorithm was very good at predicting which storms would be a class zero and cause no damage, and which storms would be at least a class three and cause a lot of damage. The researchers are adding more data for storms into the model to help improve the ability to tell class one and two storms apart from each other, in order to make the prediction tools even more useful to the energy companies.

"Our next step is to try to refine the model so it works for more weather than just summer storms," explains Tervo. "As we all know, there can be big storms in winter in Finland, but they work differently to summer storms so we need different methods to predict their potential damage."

Read the full article, Short-Term Prediction of Electricity Outages Caused by Convective Storms, by R. Tervo, J. Karjalainen and A. Jung, in the journal IEEE Transactions on Geoscience and Remote Sensing 

Thumbnail: vectora|123rf


Share Your Thoughts
Sign In or Register to leave a comment
Back to R & D Back to Top