Parameter optimization for linear power - speed control of onshore and offshore wind turbines

Τατούλης, Πέτρος; Tatoulis, Petros

dc.contributor.author	Τατούλης, Πέτρος	el
dc.contributor.author	Tatoulis, Petros	en
dc.date.accessioned	2018-09-14T08:53:50Z
dc.date.available	2018-09-14T08:53:50Z
dc.date.issued	2018-09-14
dc.identifier.uri	https://dspace.lib.ntua.gr/xmlui/handle/123456789/47586
dc.identifier.uri	http://dx.doi.org/10.26240/heal.ntua.15867
dc.rights	Αναφορά Δημιουργού-Μη Εμπορική Χρήση-Όχι Παράγωγα Έργα 3.0 Ελλάδα	*
dc.rights.uri	http://creativecommons.org/licenses/by-nc-nd/3.0/gr/	*
dc.subject	Ανεμογεννήτριες	el
dc.subject	Έλεγχος	el
dc.subject	Βελτιστοποίηση	el
dc.subject	Αεροελαστικότητα	el
dc.subject	Αιολική ενέργεια	el
dc.subject	Wind turbines	en
dc.subject	Control	en
dc.subject	Optimization	el
dc.subject	Aeroelasticity	el
dc.subject	Wind energy	el
dc.title	Parameter optimization for linear power - speed control of onshore and offshore wind turbines	en
dc.title	Βελτιστοποίηση παραμέτρων γραμμικού ελέγχου ισχύος - στροφών επίγειων και πλωτών ανεμογεννητριών	el
heal.type	bachelorThesis
heal.classification	Αιολική ενέργεια	el
heal.classification	Wind energy	en
heal.language	en
heal.access	free
heal.recordProvider	ntua	el
heal.publicationDate	2018-07-18
heal.abstract	The aim of the present thesis is the development and assessment of a calibration method for the parameters required for the linear power – speed control of both onshore and offshore wind turbines (WT). The methodology proposed falls into the general category of constrained optimization and more specifically of surrogate modeling techniques. As such, it comes with certain advantages as well as drawbacks, which will be investigated in the present work. This particular method will be referred to as Surrogate Model for Controller Tuning or SMCT throughout the thesis. The study was conducted using hGAST [1], a numerical tool developed by NTUA researchers performing the modeling and hydro-servo-aero-elastic analysis of the WT system. The specific model of WT used was the pitch regulated/variable speed DTU 10MW [2] in typical onshore, as well as in offshore floating configuration. The aforementioned WT was equipped with the Basic DTU Wind Energy controller [3], namely a double-input single-output PI controller with gain-scheduling capabilities governing the collective blade pitch angle when operating under full load conditions. Chapter 1 of the present thesis delineates the different components that comprise the methodology. As a member of the class of surrogate modeling techniques, SMCT manages to approximate the objective function chosen by the user with the highest possible fidelity within a specific design variable space. The first step is the precise sampling of design points and the calculation of the objective function’s value at these points. Subsequently, a second order regression function is produced, approximating the true objective function, thus enabling the designation of the local (or global) minimum. It is apparent that the decision regarding the number of design variables and objective functions lies fully with the user’s needs. On the contrary, the selection of the design variable domain boundaries should be handled with caution; otherwise, instability or unacceptable results may occur. The second chapter concerns the implementation of the SMCT to the onshore WT case. At first, the aerodynamic theory used to determine the controller gains and established by the DTU [4] is thoroughly presented. This particular approximate method offers an analytical expression of the proportional and integral gain with respect to the WT’s structural and operational characteristics (e.g. blades’ moment of inertia, generator mode etc.). Consequently, the user is able to crudely, but rather quickly, tune the PI controller. As a result, this method will be referred to as the Quick-Tune method. Finally, the Quick-Tune method is implemented for the reference WT and the gain-scheduling scheme is obtained. Chapter 3 includes the results of the implementation of the SMCT to the onshore WT configuration. At first, the set up of the SMCT is discussed and thereafter the SMCT, Baseline and Quick-Tune controllers are compared on the basis of WT operation stability. Moreover, the complete set of results of each one of the three optimization levels are presented along with the final tuning scheme. Some additional time series of important operating parameters are also shown, thus providing a spherical view of the behavior of the WT and the PI controller. The fourth chapter presents the results of the application of the SMCT to the offshore WT configuration. It is worth mentioning that the absence of a pre-existing tuned controller for the offshore WT divests us of the opportunity to make a detailed comparison of the operational aspects that each tuning method would result in. Thus, the tuning parameters, gain-scheduling functions and performance characteristics are demonstrated and discussed. Such performance characteristics unique to the offshore floating case include the displacement and rotation of the floater. In the framework of the present study, the set of design parameters used comprises the proportional and integral gain of the PI controller. The performance criterion used is the stability of the operation of the WT. In engineering terms, a fairly reliable indicator of stable performance is the standard deviation of the rotor speed.	en
heal.advisorName	Ριζιώτης, Βασίλειος	el
heal.committeeMemberName	Ριζιώτης, Βασίλειος	el
heal.committeeMemberName	Βουτσινάς, Σπυρίδων	el
heal.committeeMemberName	Ζερβός, Αρθούρος	el
heal.academicPublisher	Εθνικό Μετσόβιο Πολυτεχνείο. Σχολή Μηχανολόγων Μηχανικών. Τομέας Ρευστών. Εργαστήριο Αεροδυναμικής	el
heal.academicPublisherID	ntua
heal.numberOfPages	87 σ.
heal.fullTextAvailability	true