Researchers use artificial intelligence to predict electricity demand

Over the past few decades, the search for more accurate ways to predict energy consumption has been a fruitless exercise for suppliers and grid managers because Most grids still rely on forecasting models that refer primarily to consumption history and weather forecasts.

Road and rail traffic data correlates closely with activity, allowing grid managers to better understand which areas of a city or town need power and which areas need less power. In tests, the AI ​​model was combined with traditional energy consumption prediction models to make accurate predictions two to six hours before energy consumption occurs.

Real-time models are also able to provide accuracy during times of crisis, such as after a natural disaster or another pandemic. Traffic and rail data will be able to quickly identify if behavior changes and divert energy to different areas of the city.

As the number of electric vehicles grows, the connection between transportation and electricity demand will become closer. This means traffic data may become even more important in predicting electricity usage.

With wind and solar flooding the national grid, fluctuations in energy supplies have become more pronounced, making the most accurate predictions of consumption crucial for grid operators to avoid power shortages or blackouts. Coupled with growing demand for energy, past forecast models may not be able to maintain high levels of accuracy.

In follow-up tests to determine whether the AI ​​model could complement traditional models, the researchers found that it only slightly improved accuracy. Currently, it appears that artificial intelligence can be embedded into other models to provide greater accuracy.

The above is the detailed content of Researchers use artificial intelligence to predict electricity demand. For more information, please follow other related articles on the PHP Chinese website!