Survey is aimed to simulate fault conditions on wind turbine equipped with double-fed induction generator (DFIG) and to analyze the behavior of WT's control system. Mathematical model of WT with DFIG was created in Matlab Simulink. That model has a variety of input parameters and their adjustment represents a complicated optimization task. Authors propose the solution of that task with the machine learning methods, known as Reinforcement Learning. Reinforcement learning allows to interact with the environment, and can learn the set of input parameters, which produces the desired behavior of the system.
Ilyin, D.; Shestopalova, T.; and Vaskov, A.
"MACHINE LEARNING BASED ALGORITHMS FOR WIND TURBINES WITH DFIG CONTROL,"
Euroasian Journal of Semiconductors Science and Engineering: Vol. 2
, Article 17.
Available at: https://uzjournals.edu.uz/semiconductors/vol2/iss3/17