Add turbine_average_velocities property to UncertainFlorisModel

[paulf81]

Add turbine average velocity property to UncertainFLorisModel

This PR adds a property to allow getting the per-turbine average velocities from an UncertainFlorisModel. This property exists for FlorisModel. To provide this capability in UncertainFLorisModel requires mirroring the logic in get_turbine_powers to go from the expanded internal FlorisModel (which includes all the unique conditions after expanding across neighboring wind directions according to the wd_std) back to the un_expanded originally requested wind directions.

Several tests are included to check the functino works as expected

Impacted areas of the software

uncertain_floris_model.py

Test results, if applicable

New tests pass

[paulf81]

The failing tests seem unrelated to this PR but I could be mistaken

[misi9170]

Yes, the failing test is fixed in other open PRs #1126 , #1129 .

[paulf81]

Yes, the failing test is fixed in other open PRs #1126 , #1129 .

Thanks! Sorry I missed that

[misi9170]

Thanks @paulf81 , this looks good to me. I've added a few things:

• rename map_turbine_powers_uncertain() to map_turbine_values_uncertain() and create an alias with a deprecation warning
• Add test for typical velocities behavior at downstream turbines (directly downstream turbine has a higher rotor average velocity using the uncertain model than the certain one)

As we discussed, I've also:

• Added a test and an error if wd_std of 0 or less is used
• Added a comment to the ApproxFlorisModel explaining the use of wd_std=1

I think this looks good. However, there are still a lot of methods on FlorisModel that are not currently available on the UncertainFlorisModel, e.g. get_turbine_ais(), sample_flow_at_points(), etc.

[misi9170]

@paulf81 , any objections/comments on the changes I made? If not, I think I can merge.

[paulf81]

@paulf81 , any objections/comments on the changes I made? If not, I think I can merge.

Yes, these look great, thank you!