rotor effective TI in Fast.farm

[mebrahiimi]

Hi everyone,

I am building a load surrogate model trained from OpenFAST simulations, with the final goal of using it inside a wind-farm controller.

For the mean inflow, I can use a rotor-effective wind-speed estimate, for example from ROSCO I&I / wind-speed estimation. However, I am less sure about how to obtain or approximate a turbine-level effective turbulence intensity for each turbine in a way that could be available to a controller.

My question is:

What controller-available signals would be reasonable proxies for the effective TI seen by each turbine (observation, or measurement) especially for a floating IEA 15 MW turbine operating around 6.5–8.5 m/s?

Possible candidates I am considering are short-window statistics such as:

• std/mean of rotor-effective wind-speed estimate
• std/mean of rotor thrust or a thrust-based proxy
• std/mean of generator power
• std/mean of rotor speed
• blade-pitch activity, although pitch is mostly inactive in this below-rated region
• nacelle / platform-motion signals for the floating case

Any suggestions or references would be appreciated.

[jjonkman]

Dear @mebrahiimi,

Perhaps this related OpenFAST Discussion is of use to you: #2107?

Best regards,