In the context of making wind energy more interesting compared to other ways of producing electricity, algorithms to extract information about the wind that interacts with the turbine have been addressed in literature with the scope of using it to improve the turbine operations. Usual wind turbine controllers do not use wind speed information coming from the anemometer on top of the nacelle or obtained from conventional methods, this measure being too imprecise and affected by the rotor very close to the instrumentation. Generator speed and torque are used instead to regulate power production. An efficient estimation of wind speed could be very useful to improve performance of wind turbine controllers, for scheduling but also for its use in feedforward control. Moreover, it could also be useful for fault diagnosis and fault-tolerant control of the wind turbine, or to optimize power production and reduce the loads on the turbine components. The wind field reaching the turbine may be very complex and difficult to characterize. In literature several approaches have been defined to extract a single quantity, the effective wind speed, representative of the incoming airflow. This thesis deals with the implementation of one of the algorithms suggested in literature on the DTU 10 MW Reference Wind Turbine numerical model, to design a wind observer to extract the effective wind speed. A wind observer is subsequently designed for the PoliMi Wind Turbine Model, a scaled version of the DTU 10 MW Reference Wind Turbine developed for wind tunnel tests at PoliMi, and tested on its numerical model. A feedforward control action that uses the information about the estimated effective wind speed is then designed to improve the standard feedback controller of the DTU 10 MW Reference Wind Turbine reducing the fatigue loads on the turbine components and decreasing the fluctuations of power, torque and rotational speed. The same concept is finally adapted to the PoliMi Wind Turbine Model.
Per rendere l'energia eolica più interessante rispetto ai modi convenzionali di produrre elettricità, in letteratura è stata considerata la possibilità di utilizzare algoritmi per estrarre informazioni sul vento che interagisce con la turbina, allo scopo di migliorarne l'utilizzo. Normalmente il controllore della turbina non tiene conto del vento misurato tramite l'anemometro in cima alla turbina o in modi simili, essendo questa misura imprecisa e influenzata dal movimento del rotore. La velocità angolare e la coppia al generatore sono invece utilizzate per regolare la produzione di potenza. Una stima efficace della velocità del vento potrebbe essere utile per migliorare il comportamento del controllore, per lo scheduling del controllo o per l'utilizzo nel controllo in feedforward. Altri possibili utilizzi sono l'individuazione di guasti e il controllo tollerante ai guasti, oppure l'ottimizzazione della produzione di potenza e la riduzione dei carichi sui componenti della turbina. Il campo di vento che raggiunge la turbina può essere molto complesso e difficile da caratterizzare. In letteratura esistono molti approcci per estrarre una singola quantità chiamata effective wind speed (velocità del vento efficace), che sia rappresentativa del flusso d'aria. Questa tesi affronta l'implementazione di uno degli algoritmi suggeriti in letteratura per estrarre la velocità del vento efficace sul modello numerico della turbina eolica DTU 10 MW Reference Wind Turbine. Successivamente l'osservatore del vento viene svilluppato per il PoliMi Wind Turbine Model, una versione in scala del DTU 10 MW Reference Wind Turbine realizzato per test in galleria del vento al PoliMi, e testato sul suo modello numerico. Viene poi realizzata una azione di controllo in feedforward che utilizza il vento stimato per migliorare il controllo standard in feedback sul DTU 10 MW Reference Wind Turbine per ridurre i carichi di fatica e le fluttuazioni di potenza, coppia e velocità angolare. Lo stesso concetto è infine adattato al PoliMi Wind Turbine Model.
Numerical design of a wind observer and feedforward control of wind turbines
CANISTRO, FEDERICO
2017/2018
Abstract
In the context of making wind energy more interesting compared to other ways of producing electricity, algorithms to extract information about the wind that interacts with the turbine have been addressed in literature with the scope of using it to improve the turbine operations. Usual wind turbine controllers do not use wind speed information coming from the anemometer on top of the nacelle or obtained from conventional methods, this measure being too imprecise and affected by the rotor very close to the instrumentation. Generator speed and torque are used instead to regulate power production. An efficient estimation of wind speed could be very useful to improve performance of wind turbine controllers, for scheduling but also for its use in feedforward control. Moreover, it could also be useful for fault diagnosis and fault-tolerant control of the wind turbine, or to optimize power production and reduce the loads on the turbine components. The wind field reaching the turbine may be very complex and difficult to characterize. In literature several approaches have been defined to extract a single quantity, the effective wind speed, representative of the incoming airflow. This thesis deals with the implementation of one of the algorithms suggested in literature on the DTU 10 MW Reference Wind Turbine numerical model, to design a wind observer to extract the effective wind speed. A wind observer is subsequently designed for the PoliMi Wind Turbine Model, a scaled version of the DTU 10 MW Reference Wind Turbine developed for wind tunnel tests at PoliMi, and tested on its numerical model. A feedforward control action that uses the information about the estimated effective wind speed is then designed to improve the standard feedback controller of the DTU 10 MW Reference Wind Turbine reducing the fatigue loads on the turbine components and decreasing the fluctuations of power, torque and rotational speed. The same concept is finally adapted to the PoliMi Wind Turbine Model.File | Dimensione | Formato | |
---|---|---|---|
2019_04_Canistro.pdf
non accessibile
Descrizione: Testo della tesi
Dimensione
22.7 MB
Formato
Adobe PDF
|
22.7 MB | Adobe PDF | Visualizza/Apri |
I documenti in POLITesi sono protetti da copyright e tutti i diritti sono riservati, salvo diversa indicazione.
https://hdl.handle.net/10589/146244