Tuning ROSCO controller with DISCON.IN

Wind & Water Controls
21

Dear all,

I have a question regarding the “floating feedback” control strategy. I’m generating some customized DISCON.IN files using the ROSCO toolbox, with each file featuring different closed-loop frequencies and damping values for the pitch controller (zeta_pc and omega_pc).

My question is:

If I set Fl_Mode: 0, zeta_pc: 1.0, and omega_pc: 0.2 in the turbine’s YAML file, I will generate a DISCON.IN file where the gains of the pitch controller are calculated to ensure the dynamic performance is as specified (zeta_pc = 1.0, omega_pc = 0.2).

However, if I set Fl_Mode: 1 while keeping the same zeta_pc and omega_pc values, will I achieve different dynamic performance? If the answer is yes, how can I calculate the new closed loop frequency and damping?

I hope my question is clear.

Best regards,
Marco

image
image1684×429 67.8 KB

22

Dear all,

I am investigating the floating feedback controller used for the IEA 15 MW turbine on the VolturnUS-S semisubmersible platform, as described in the NREL reference report.

From Section 2.4 of the report, the tower fore-aft motion is fed back into the pitch controller through a proportional gain (reported as approximately −9.32 using ROSCO tuning logic). My understanding is that the negative sign of this gain is intended to increase platform-motion damping.

In my study, I implemented the floating feedback together with the fixed-base controller bandwidth, i.e., a pitch control natural frequency of 0.2 rad/s (≈ 0.032 Hz) with FL_mode=2. I compared this against a detuned controller configuration with a reduced bandwidth of 0.0628 rad/s (≈ 0.01 Hz), without floating feedback.

The simulations were carried out under IEC design load conditions (NTM), using wind speeds in the range of 12–16 m/s and the corresponding wave parameters (significant wave height and peak period) from the reference table in the report.

From my results, I observe that the floating-feedback controller (with the higher bandwidth of 0.2 rad/s) provides improved performance compared to the detuned controller in terms of generator speed, power, and most platform motions (except surge).

However, this seems counterintuitive to me. Based on the literature, I would expect that a damping-oriented floating feedback (negative gain), especially when combined with a higher-bandwidth pitch controller, would increase generator–platform coupling and therefore lead to larger generator-speed and power variations compared to a detuned controller.

Therefore, I would like to ask:

  1. Is my understanding correct that the negative floating-feedback gain is primarily intended for platform damping rather than coupling reduction?

  2. Why does the floating-feedback controller (with higher bandwidth) still show improved generator-speed and power responses compared to a detuned, lower-bandwidth controller?

  3. Is the observed improvement mainly due to increased aerodynamic damping from the higher-bandwidth pitch controller, rather than a reduction in generator–platform coupling?

  4. What behavior should be expected if a positive floating-feedback gain is applied together with the fixed-base bandwidth controller, particularly in terms of platform stability and generator dynamics?

Any insights or clarification would be greatly appreciated.

Thank you.

23

Hi Nitin,

Your understanding is correct.

My interpretation is that since the platform pitch is damped, it reduces the rotor motion and by extension, the coupling is less impactful.

A nice article detailing what you are asking about can be found here: https://onlinelibrary.wiley.com/doi/10.1002/we.2864

Usually, when a positive floating-feedback gain is applied, the performance is much worse overall.

I hope this helps.

Best, Dan

1 Like