Real component of PTO impedance must be constant for all frequencies Error

[degoeden]

Hi,

I was trying to create a PTO with an impedance that has a real component that is dependent on frequency when I ran into this error. Upon reading the error message and looking in the code, it makes a lot of sense that I am receiving the error, but I'm confused as to why this check is present. It seems to limit my PTO circuit design to only use components in series so that I can use KVL (I was trying to model a circuit that includes elements in parallel which forces me to use KCL, which forces me to have real components that depend on frequency).

I disabled the check for now (unfortunately I can't say if it works or not yet because I have been running into another error) but was wondering if you could tell me why this check is there, and if it will be problematic that I disabled it. Thank you!

[dtgaebe]

Hi Nate,

I think for now it is good that you disabled the check. If I remember correctly we implemented the check because we anticipated only PTO impedances with constant loss (real part) due the linear nature.
However, we very well might be wrong with that assumption.

Do you want to share more about the PTO you're trying to model? Feel free to do this offline.

[rgcoe]

For reference, here's the check that was giving @degoeden trouble. The decision to include this related to how we assign the f=0 elements of the PTO impedance (we set $\Re \{ Z_{pto} (f=0) \} = \Re \{ Z_{pto} (f_1) \}$, and thought that real part of the PTO impedance should indeed be constant... current thinking is that this check should maybe be removed and that the real part of the PTO impedance is not necessarily constant.

WecOptTool/wecopttool/pto.py

Lines 137 to 144 in 528c679

	check_3 = (
	np.allclose(np.real(impedance[:, :, i+1]), np.real(impedance[:, :, 0]))
	)
	if not check_3:
	raise ValueError(
	"Real component of impedance must be constant for " +
	"all frequencies."
	)

[dtgaebe]

To cross-reference related issues:
#309 (comment)
and #225

Long story short: The PTO impedance at f=0 makes the optimization problem well defined for the DC component of unstructured controller when optimizing for electrical power