A software upgrade developed by researchers at France's Polytechnic Institute of Paris improves the efficiency of wind turbines by ensuring they spend more time facing directly into the wind.
The researchers accomplished this by training a reinforcement-learning algorithm to track wind patterns and formulate a strategy to keep the turbine facing the appropriate angle.
The current strategy of adjusting the turbine blades in accordance with wind patterns uses more energy and causes wear and tear on the components.
In simulations, the new algorithm was more efficient than the current algorithm, reducing the amount of time spent readjusting turbine positions to 3.7% while generating power gains of 0.4%.
The new model could increase electricity production by 5 terawatt hours per year if implemented globally.
From New Scientist
View Full Article - May Require Paid Subscription
Abstracts Copyright © 2023 SmithBucklin, Washington, DC, USA
No entries found