Good evening,

when analyzing the turbine start-up I discovered a anomaly comparing the control inputs (e.g. generator torque/power and blade pitch angles).

For some reason the generator torque from the actuator (compare fig1, “Out1”) deviates from the processed value at the output (in “OutData”).

[attachment=1]FAST_v8_Ctrl_150831_BC.png[/attachment]

In Fig.2 you can find an examplary plot of the generator power.

[attachment=0]150903_Power_Comparison_bri_sma.png[/attachment]

After approximately 0.1 seconds both input and output converge to almost exactly the same value which is true for the rest of the 600 seconds simulation. So this problem appears only at starting up FAST.

Unfortunately, I am interested in these first 7 to 9 time steps to match exactly. This is because my model predictive controller can only calculate the control inputs outside of simulink and therefore I have to run FAST always only for one sample time and then compute new control input values.

Hence, the difference between the blue line (my commanded control input) and the green line (used FAST 8 value) produces rather big issues controlling the wind turbine at all. (Sometimes similiar anomalies appear for the commanded pitch angles but I have not yet figured out, if this is a reproducable phenomenon.)

Does someone know a solution for the issue or encountered the same described problem? And why does both values deviate from another although they are control inputs (which actually should not be adjusted by FAST)?

Thanks in advance for your replies!

Kindest Regards,

Bastian Ritter