Page 1 of 1

Linearization of the full system

Posted: Mon Dec 03, 2018 6:45 pm
by Guo.Wu
Hello Everybody,

I am working on a kind of floating wind turbine platform by openfast. I should work out the linearization model for the design of control DLL. But maybe openfast can not cover linearization of the full system at this stage from the topic "The frequency component of flapwise". That means hydordyn, mooring modules can not be linearized by openfast now.

I want to ask If some other approximate substituted solution exists?

Re: Linearization of the full system

Posted: Mon Dec 03, 2018 8:36 pm
by Jason.Jonkman
Dear Guo.Wu,

There are generic system identification approaches to approximate linear state-space models, but a direct linearization would be more accurate. I think you'll have to wait for us to release the new linearization functionality in OpenFAST for floating wind turbines.

Best regards,

Re: Linearization of the full system

Posted: Tue Dec 04, 2018 2:46 am
by Guo.Wu
Dear Dr. Jonkman,

Many thanks for your fast reply, I will consider your solution , and I am looking forward to the update of openfast.

Re: Linearization of the full system

Posted: Tue Dec 04, 2018 6:40 pm
by Guo.Wu
Dear Dr. Jonkman,

Another approximative method I am thinking about is that. An additional section will be added at the bottom of the tower. The length and stiffness of the added section will be adjusted to make sure the mean displacement of the tower bottom will as same as the displacement when the floating platform added at the bottom. The Linearization of the full system will be approximated based on the adjusted model by openfast at this stage. I want to ask if this is a reasonable method?

Re: Linearization of the full system

Posted: Tue Dec 04, 2018 7:49 pm
by Jason.Jonkman
Dear Guo.Wu,

You intend to do this through the adjustment of the stiffness of the tower near its base? Due to the potentially large motion of the platform, I would think it would difficult to represent the floating-platform motion as a tower mode shape described by a low-order polynomial.

Best regards,

Re: Linearization of the full system

Posted: Wed Dec 05, 2018 12:23 am
by Guo.Wu
Dear Dr. Jonkman,

Thank you very much for your advice.

Re: Linearization of the full system

Posted: Fri May 03, 2019 9:22 am
by zarei.atefe
hello
i have a question about linearization.I want to design a PID controller for pitch control. I need to use linearization and must pcmode = 0 in the linearization, but in this mode, the pitch control command is not taken from the simulink and the pitch angles are commanded zero.
How can I use pcmode = 4 and linearization ?
thanks.

Re: Linearization of the full system

Posted: Fri May 03, 2019 9:53 am
by Jason.Jonkman
Dear Zarei,

If you want to design a PID controller, presumably you don't already have the controller. A linearization with PCMode = 0 simply means that the controller logic is not linearized and included in the state-space model. The pitch angles need not be zero, but the operating point pitch angles will be fixed at whatever you set in your FAST / OpenFAST model when linearizing. The inputs to the controller (outputs from various modules) and the controller commands (e.g. perturbations of the blade-pitch angle about the operating point) still show up as inputs/outputs in the linear state-space model, from which you can design your PID controller.

Best regards,