Model predictive control of Type 4 wind turbines
Author
Kristiansen, Rasmus Løvschall
Term
4. term
Education
Publication year
2023
Pages
125
Abstract
Specialet undersøger modelprædiktiv styring (MPC) som converter-kontrolstrategi for Type 4 vindmøller. Der udvikles en prædiktionsmodel, der indfanger de relevante systemdynamikker, og MPC’en formuleres som et kvadratisk optimeringsproblem med lineære begrænsninger, der afgrænser både outputs og kontrolsignaler. Controlleren implementeres i C med henblik på indlejret hardware og evalueres i et high-fidelity simuleringsmiljø under normal drift, netfejl og variationer i grid-styrke. Testresultaterne viser stabil og præcis regulering, pålidelig håndtering af netfejl og god robusthed, hvilket indikerer at MPC er en anvendelig strategi til converter-kontrol i Type 4 vindmøller.
This thesis explores model predictive control (MPC) as a converter control strategy for Type 4 wind turbines. A prediction model capturing the relevant system dynamics is developed and used to formulate the MPC as a quadratic optimization problem with linear constraints that bound both outputs and control signals. The controller is implemented in C to support embedded hardware and evaluated in a high-fidelity simulation environment under normal operation, grid fault events, and variations in grid strength. Test results show stable and accurate regulation, reliable fault handling, and strong robustness, indicating that MPC is a viable strategy for converter control in Type 4 wind turbines.
[This summary has been generated with the help of AI directly from the project (PDF)]
Keywords
Documents