Wind turbines that use human-like learning to improve efficiencyView gallery - 2 images
Wind turbines are exposed to a wide variety of wind conditions, from zephyrs to gales, and ensuring the maximum amount of power is extracted from the turbine across a range of wind speeds is a difficult task. Chinese researchers have now developed a biologically inspired control system that uses “memory” of past experience to learn how to best adapt to changing conditions.
Wind turbines are designed with a rated power and a rated speed, which is the wind speed at which the turbine will produce its rated power. For example, a 10 kW wind turbine with the most common rated speed of 25 to 35 mph (40 to 56 km/h) will only generate the designated 10 kW at those speeds.
Sick of Ads?
More than 700 New Atlas Plus subscribers read our newsletter and website without ads.
Join them for just US$19 a year.More Information
When the wind falls below or exceeds this ideal range, control systems kick in to alter the turbine system to help keep power efficiency high in low winds and protect the turbine from damage in high winds. These changes can include altering the angle of the blades, modifying the electromagnetic torque of the generator.
Turbine control systems are comprised of three basic elements – sensors to gather data, actuators to carry out changes to the turbine system, and algorithms to coordinate the actuators based on the data supplied by the sensors. While these control system algorithms often rely on complex computation models of the turbine’s behavior, a group of Chinese researchers have developed a control system inspired by human learning models.
The new control system, which is described in the Journal of Renewable and Sustainable Energy uses memory of past control experiences and their results to generate new actions. While simulations of the system produced poor initial results, it quickly learned how to improve to match the performance of a traditional control system, while being much simpler.
The researchers claim their “human-memory-based method holds great promise for enhancing the efficiency of wind power conversion.”View gallery - 2 images