Improved control could optimize wind power

Nov. 11, 2019
Researchers consider long-range generalized predictive control theory as an alternative to direct power control for wind power

Although wind power has potential, it can be expensive to operate and maintain, and often can only process about 30% of the generated power, according to Juan Sebastián Solís-Chaves, professor with the Department of Energy Engineering at the Federal University of ABC-UFABC in Brazil. In an effort to optimize this process, researchers in Brazil analyzed the commonly used double-fed induction generator-based configuration and considered non-linear control theory as an alternative.

The researchers recently published the results of the study in IEEE/CAA Journal of Automatica Sinica, a joint publication of IEEE and the Chinese Association of Automation.

The traditional system consists of two windings, or electromagnets, that feed information from the outside environment to the generator. They can adjust the frequency at which power is generated based on the strength of the wind spinning the turbine. However, the double-powered system requires a converter to change the power from alternating current to direct current, depending on what speed conversion is needed. For larger systems, this can be a significant expense.

“Of all possible arrangements for wind energy systems, the most commonly used is the double-fed induction generator-based configuration,” paper author Solís-Chaves explains in a statement. “We set out to investigate how long-range generalized predicative control theory could be an interesting and innovative alternative to the traditional direct power control used in wind energy systems.”

In the traditional generator, the controls address the entire system together. Despite having different stages, each stage is dependent upon the others, and must be considered together. For example, if the wind strength increases and spins the turbine more, the system will adjust the converter and power generation processes together, using each as a variable for the others.

In their analysis, the researchers separated the system to treat power generation and feedback as their own systems with their own input and output. "This is a non-linear control theory that is a very robust alternative with a faster dynamic response than traditional systems," Solís-Chaves said.

In this wind energy system using the researchers new control system, the turbine can more quickly respond, lessening the strain on its physical components. The researchers plan to continue fine-tuning their de-coupled generation system.