The Chinese Invented Wind Generator with Memory
Engineers have developed a new method of control for the wind turbine that can increase the performance of the installation. According to Leonid Popov, the method was inspired by human memory and ability to learn.
Any wind generator needs the performance adjustment, which is necessary for maximum efficiency at low wind speeds as well as for the prevention of rotor damages at strong flows. The adjustment is made by changing the angle of the attack of the blades, due to the momentum electromagnetic correction on the shaft of the generator, and, finally, owing to the braking system switching.
Usually, when making the control program, a complex digital model of the turbine performance is entered. As the dynamics of the rotor under the conditions of chaotic external disturbances are not linear at all, this method results in great inaccuracies in the control activities. Thus, the unit produces less energy than it could.
Chinese researchers proposed another option. According to EurekAlert, in their system, the turbine remembers its previous control experience and uses it to assess the current situation and identify new activities. Thus, there is no need in a highly complex model of the rotor response to certain environmental influences. The system just learns itself on the fly.
Still it was a kind of simulation. However, during such numerical modelling, at first the memory controller showed poor results, but soon it “understood” the best way of activities. It received very high return from the generator when the wind level was lower than the nominal one for the calculated windmill, and it held the nominal power steadily when the wind level was more than the calculated one.
You can find the detailed research in the Journal of Renewable and Sustainable Energy article.
Also you can read:
