The efficiency of wind parks is limited by the fact that the movement and force of the wind is impossible to predict. Google, and the artificial intelligence company DeepMind, which is also owned by Google, is aiming to make it possible: last year, the two companies have started testing machine learning on Google’s own wind turbines.
The results are promising: DeepMind’s machine learning platform made wind power prediction 36 hours ahead, which allowed Google to make supply commitments to power grids in advance. The method has boosted the value of the wind energy produced by Google’s wind turbines, compared to the baseline scenario of no time-based commitments to the grid, DeepMind wrote in a statement.
Source: Engadget