Low-Rank Distributed Control with Application to Wind Energy

This thesis addresses three different topics in wind power plant operation. Most of the research is focused on controlling a wind farm that is required to meet a power set-point. In this mode of operation, the wind turbines are able to vary their power production as long as the total power demand is met. The research investigates how this freedom can be used in order to reduce the fatigue loads experienced by the wind turbines. The problem is studied in a linear-quadratic control setting where the objective is to minimize the total fatigue load experienced by the turbines, while satisfying a requirement on their joint power production. It is shown that, under certain assumptions, the design problem can be drastically simplified. In particular, the computational effort needed to obtain the solution is independent of the number of wind turbines and the only centralized operation required to implement the optimal control law is a single summation. The research also explains the mechanisms that make power allocation schemes useful for load reduction. Part of the research addresses wake effects in wind farms by developing a low-complexity model of the aerodynamic interaction between wind turbines. The model is used in a series of examples, where the wind turbines coordinate their power productions in order to maximize the power production of the wind farm. The examples indicate that the benefit of power coordination increases with the number of turbines in the wind farm. They also identify the underlying mechanisms behind this effect. The last topic of the thesis is to investigate the benefits of using preview of the incoming wind speed in order to reduce structural loads on the wind turbine tower. The main focus is to understand how measurement distortion influences the achievable load reduction as well as the required length of preview. Results from high-fidelity simulations based on real wind turbine measurements indicate that the use of preview can lead to a significant reduction of tower fatigue loads and that the length of preview needed to attain the reduction does not exceed a few seconds.

Grantor dc:publisher

Department of Automatic Control, Lund Institute of Technology, Lund University

Year dc:date

2014

Author dc:creator

• Madjidian, Daria

• Control Engineering

Statement dc:rights

• info:eu-repo/semantics/openAccess

Language dc:language

eng

Identifier

urn:isbn:978-91-7623-004-6 (web)
urn:isbn:978-91-7623-003-9 (print)

OAI identifier oai:identifier

oai:lup.lub.lu.se:9ff63e87-e754-414d-a21f-275cf2c074b6