Blowing Past the Competition

EPFL researchers used a genetic algorithm to enhance vertical-axis wind turbine design, improving efficiency and mitigating stalling issues.

Nick Bild
28 days ago β€’ Machine Learning & AI
An experimental vertical-axis wind turbine (πŸ“·: Alain Herzog CC BY SA)

Harnessing wind power to generate electricity is one of the most promising sources of renewable energy available today. The principle behind wind power is simple: turbines capture the kinetic energy of wind and convert it into electrical energy. This process is clean, sustainable, and has minimal impact on the environment compared to fossil fuel-based energy generation methods.

Unlike traditional power plants that rely on burning fossil fuels, wind turbines produce electricity without emitting harmful pollutants or greenhouse gasses. This makes wind power a crucial component in reducing air pollution, contributing to a cleaner and healthier environment.

Furthermore, wind is a widely available resource, with potential for harnessing it present in many regions across the globe. This decentralization of wind resources means that communities can produce their own electricity locally, reducing reliance on centralized power grids and enhancing energy security.

But as is the case with any technology, wind energy harvesting technologies have their limitations. The windmill-style horizontal-axis wind turbines that are widely used for energy production in wind farms today alter wind conditions and decrease the effectiveness of downwind turbines. This limits the amount of exploitable land that is available in any particular region. Moreover, these turbines also have a nasty habit of harming wildlife with their spinning blades.

These issues could be resolved by revisiting an old technology β€” vertical-axis wind turbines. These turbines spin perpendicular to the wind and spin more slowly, yet have a greater energy density. Adding different turbine styles into the mix could help to avoid undesirable alterations of wind patterns, and the lateral rotation of the blades is easier for wildlife to avoid. As an added bonus, vertical-axis wind turbines are also quieter.

Yet these vertical-axis turbines are rarely used in practice due to their aerodynamic complexity. The blades encounter varying airflow conditions in the course of a single rotation, even in steady wind conditions, which can cause them to stall. Gusts of wind serve to further amplify this issue. But that problem may soon be in the rearview mirror, thanks to the efforts of a pair of researchers at the Swiss Federal Institute of Technology Lausanne. They have leveraged machine learning techniques to help them design a better vertical-axis turbine that does not suffer from issues with stalling.

The team equipped an existing vertical-axis wind turbine with sensors to measure the forces placed on it by the wind. Data was then captured as they adjusted the pitch of the blades, as well as their speed and amplitude. This information was then fed into a genetic algorithm that ran through thousands of experimental iterations in which it was seeking out the optimal values for the adjustment of the turbine blades.

Using these results, the researchers identified two parameter profiles that significantly enhanced both the efficiency and robustness of vertical-axis turbines. It was found that this new design actually leveraged the forces that normally cause a stall to push the blades forward, eliminating a major issue with current designs.

To date, the design has only been tested under controlled laboratory conditions. A plan is in place to build another proof of concept device that will be tested outdoors, however, where its real-world performance can be assessed. If it proves to work as well as early testing indicates, the researchers hope to commercialize their technology.

Nick Bild
R&D, creativity, and building the next big thing you never knew you wanted are my specialties.
Latest articles
Sponsored articles
Related articles
Latest articles
Read more
Related articles