Can you kindly help me identify how can I get the state space matrix representation for a linearized wind turbine model resulting into A,B,C and D matrices for different operating points (i,e different wind speeds) ?
I want to use these state space matrices for pitch control of a wind turbine system. Also for simplicity I am looking for a no bigger then a 3rd order model or 3 DOF mode.
(Note: Every point will be generating its own set of A,B,C and D matrices)
Thank you & Obliged,
Ibrahim
Dear Somesh,
Thank you for your response. I am looking for a SISO model. Taking pitch angle as its input and and giving a constant generator speed as the output. Operating in the region 3 of the wind turbine (see attached paper) giving me a rated power as my output for the specified region in which I plan to design my controller for. The paper I am attaching here gives me a State Space linearized model for about one operating point (i.e wind speed 12 m/s), giving me A,B,C,D and other linearized matrices.
What I am looking for is, the same A,B,C & D matrices for different wind speeds (not just about one point, as specified in the paper) for any wind turbine. I couldn`t find it in my literature survey so far, I was able to identify a software FAST which can be used. I was hoping if someone could share his/her results for the linearized state space model with me for the different wind speeds (operating point) and not just about one operating point ( as most of the papers I come across gives me these state matrices about a single point, for the horizontal axis wind turbines).
Thank you & good day.
The aerodynamic of a wind turbine is strongly nonlinear, due to the product between omega2 and incidence (which mainly depends on pitch and wind speed). Then you absolutely need to linearize your system to obtain a state space representation. How to do that depends on the way you build your model. If you have a set of equation you must substitute each variable x with the sum of its equilibrium value x0 and a perturbative part x' . Then you impose the equilibrium condition and take the limit for x' going to 0. You can Google for linearizarion
You can actually look at my thesis, I did some similar work for pitch control with DFIG, using linearization for adaptive NN.
http://search.proquest.com/docview/1313228891
Hope this helps you.
Ibrahim,
You are looking for an extremely simplified model for a very complicated system - to the extent that on today's date your requirement will be termed as trivial !
An old publication that could help you is available at:
http://www.nrel.gov/docs/fy99osti/26244.pdf
But the work is obsolete, if anything ! See if this will do ?
With best wishes.
-Sanjay
From the information you give, it seems you you need to start from a nonlinear system in order to linearize it around the operatng points you want. I recommend you to read about the Lyapunov's direct method in the book written by Slotine and Li, "Applied Nonlinear Control". I think that will solve your problem instead of the paper you give above because that one considers the H-infinity approach and might be difficult to start with.
- Badges
- Science topic
- Similar topics
- Engineering
- Modeling