As the size of wind turbines steadily increase, a control system which can manage the loads and dynamics becomes more important. In this thesis, the effects of turbine scale on the control system are examined and designs which mitigate the arising problems are presented and discussed.In this thesis, a set of three wind turbines is developed using a method to scale a mathematical model of a wind turbine while maintaining similarity in the dynamics. This framework for producing the scaled wind turbines is presented and discussed.The performance of the controller for a very large wind turbine is limited by the dynamics of the tower. By accounting for the non-minimum-phase dynamics present in the wind turbine, previous work has reduced loads in the tower. In this thesis, this framework is developed to improve speed and power control and recover some of the performance lost as turbine size increases.As well as the effect of the tower, non-linear dynamics present in the pitch control loop adversely effect performance. Previous work has developed a framework for a controller for non-linear plants which satisfies a criteria called extended local linear equivalence (ELLE). A novel controller which satisfies the ELLE criteria is presented which counters the non-linear dynamics present in the wind turbine and reduces fluctuations in speed and power.A comparison of a baseline controller and a controller which incorporates the two designs described above shows significant reductions in the fluctuations of rotor and generator speed as well as power output. These changes to the controller also show greater improvements to performance in larger turbines. The inuence of the the tower and the non-linear dynamics present in the aerodynamics both become more severe as the size of the wind turbine increases. Therefore, a controller design which mitigates these effects has greater value as the wind energy industry continues on its path and develops ever larger wind turbines.
|Date of Award||1 Mar 2014|
- University Of Strathclyde
|Sponsors||University of Durham|
|Supervisor||William Leithead (Supervisor) & Hong Yue (Supervisor)|