Virtual load sensing for wind turbine gearboxes

This PhD will explore the merits of virtual sensing for load estimation in wind turbine drivetrains. The PhD consists of three main parts: Modelling, Virtual Sensing and Validation.

In the modelling part, wind turbine drivetrain models of increasing levels of complexity will be generated. The complexity is varied in two directions. On one hand the accuracy of modelling of different subsystems is increased. On the other hand the scope of the model is broadened to include, in addition to the gearbox, the rotor, generator, …

The virtual sensing part will investigate the accuracy and bandwidth that can be achieved with virtual load sensing. With virtual sensing the previously generated models are combined with live measurements optimally using state estimation tools. The link between modelling complexity and performance of the virtual sensor will be studied.

The validation part will verify the performance of the virtual sensors on a number of different test scenarios.