Background

Before being implemented on real wind turbines, all control algorithms have to be thoroughly validated. Besides aeroelastic simulations, also wind tunnel experiments with scaled models are used for such purposes at the University of Oldenburg. The available setup consists of a scaled wind turbine model MoWiTO 1.8 placed in the open section of the wind tunnel. The active grid positioned at the wind tunnel inlet allows for reproducible generation of different inflow conditions, such as turbulent wind, gusts, and wakes.

Typically, mathematical models used for the development of wind turbine control do not fully represent the wind turbine behaviour, especially during its entire lifetime. To guarantee desired performance and stability despite imperfectly captured nonlinear phenomena and uncertainties inevitably present when using  reduced-order models, robust and fault-tolerant control as well as approaches for fault detection and isolation are  promising approaches for enhancing the system behaviour and reliability.

Task

The main task is to develop a robust controller for the scaled wind turbine MoWiTO 1.8 which can guarantee stability and desired performance in a wide range of inflow conditions. To achieve this aim, already available experimental data will be  used to identify structured and unstructured uncertainties in the mathematical model of MoWiTO 1.8. After designing and validating the developed controllers in simulations, selected approaches will be implemented in LabVIEW and validated in wind tunnel experiments.

Work steps

• Literature review on robust and fault-tolerant control as well as wind turbine modeling
• Uncertainty quantification based on experimental data
• Design and simulation-based verification of the controller
• Implementation on the available control hardware (NI cRIO)
• Validation of the controller in the wind tunnel
• Documenting and presenting the work

Requirements

• Interest in wind turbine control
• Interest in experimental work
• Basic understanding of wind turbines and control theory
• Basic experience or willingness to learn LabVIEW