Improving efficiency of Solar Panels

Tracking sun with Artificial Intelligence

Solar panels that track the sun increase energy production by 15% to 40% depending on the season and location. The common solar tracking algorithms usually estimate the position of the sun and point the panels into that direction.

However, a variety of factors contribute to the performance of a panel and pointing it directly at the sun is not always the best strategy. For example, atmospheric conditions, clouds, shades, foliage and reflectance from the environment (like snow) may affect the optimal position. By using deep reinforcement learning we can teach AI to figure out the optimal position in the matter of milliseconds and adjust the panels accordingly.

The total solar irradiance falling on a panel is a combination of direct, reflective, and diffuse irradiance
The reflective irradiance varies heavily based on the percentage of irradiance reflected off the surrounding ground surface
Classical tracking approaches require additional hardware such as a barometer, thermometer, or GPS
AI methods outperform traditional algorithms like Grena tracker

Sun Tracking with AI

Artificial Intelligence powered solar panel controller can account for weather change, power use, shading effects, cloud coverage, and diverse reflective indices of surroundings, offering an efficient and adaptive solution that can optimize for the given environmental conditions without the need for complex hardware, regardless of the location or time of the year.

See Tableau Workbook