09:00 - 10:30 Control & LiDAR
Novel approaches to balancing competing and multivariable control objectives, including results from field tests
- LiDAR-assisted feedforward control algorithms
- An adaptive data processing technique for LiDAR-assisted control
- Advanced H∞ controllers
- Delegates will be able to list LiDAR technologies
- Delegates will be able to identify how LiDAR-assisted control can be developed, independent of turbine
- Delegates will be able to explain alternative approaches to designing multivariable control
Lead Session Chair:
Mauro Villanueva, Gamesa, Spain
Jan van Wingerden, TU Delft, The Netherlands
Pablo Vital (1) F Octavio Hernandez (2) Alberto Moreno (3) Jaime Suarez (4)
(1) Gamesa, Sarriguren, Spain (2) Adwen Offshore, Madrid, Spain (3) Matis Groupe, Madrid, Spain (4) Gamesa, Madrid, Spain
Printer friendly version: Print
Presenter's biographyBiographies are supplied directly by presenters at EWEA 2015 and are published here unedited
Mr. Pablo Vital is a senior control engineer of the Technological Development department at Gamesa working on researching projects. He has been involved in the Control System field over the last 17 years and in the wind industry since 2001, first at MTorres company during 9 years and after at Gamesa. In these years, he has participated in a wide range of wind projects, from field tasks like operations and maintenance, assembly, commissioning to certification processes. His area of expertise is the modelling and control of dynamic systems and the development of innovative wind energy control solutions.
Yaw Offset Correction based on Extremum Seeking Control
Yaw offset is referred to any circumstance causing a wrong tracking of the wind turbine alignment. A direct consequence of this effect is a lower power curve and an increase in structural loads. Possible causes of yaw offset are wrong sensor assembly or inadequate parameter adjustment. Nevertheless, the aerodynamic influence of the rotor over the wind stream downwards is probably the most relevant problem. This influence might cause misalignments higher than the limits established by the main guideline and normative standards.
Traditional solutions involve additional hardware, i.e. sensor placed in front of the rotor, mast or lidar measurements, etc… but obviously, this has an undesirable economic impact.
The proposal of this work is to present a methodology which corrects any type of yaw offset based on mathematical algorithms implemented in the wind turbine control software. For this purpose, Extremum Seeking Control (ESC) techniques have been investigated and adapted to develop an optimization algorithm.
Specifically, ESC is a maximum power tracking algorithm based on a gradient method. It works introducing a slight perturbation over the variable to be optimised, such that a perturbation at the same frequency is provoked over the cost function to be maximized. In that case, the cost function is the power ratio (similar to the power coefficient) and the variable to be optimised is the yaw offset. Hence, an imposed movement over the yaw system will cause a variation on the global efficiency allowing the algorithm to converge to the best working point.
Main body of abstract
The final scheme has been developed to work on an automatic and unattended mode, full-compatible with normal working of the wind turbine, including on-off transients, alarms or maintenance actions. This scheme together with the algorithm characteristics provide a rapid and secure convergence.
Field test have revealed good accuracy and robust performance, but also an important practical issue to be necessarily taken into account: yaw misalignment seriously affects speed and direction measurements of the nacelle sensors. This effect is not detected at simulation stage. For this reason, an important iterative work has to be done to cope with it.
Furthermore, the use of the algorithm on field suggests the possibility of determining a set of yaw offsets as a function of wind speed and direction sector.
Future work to be done is to get a better understanding of the rotor influence over the wind stream downwards, both in direction and speed, as a function of the yaw offset itself.
As a conclusion, a software solution is proposed and tested on a real wind turbine to solve yaw offset problem in a practical way. Field tests have shown excellent performance, but also presenting some difficulties due to the aerodynamic influence of the rotor over the wind stream downwards.
Extremum Seeking Control techniques have made possible an accelerated and secure convergence to effectively solve this complex problem. The use of this theory comes up as a very promising method to improve other problematical issues like fine pitch tracking or optimum torque in wind turbine region 1.