search

byuflowlab / Solar.jl

Package Containing Methods related to solar energy collection.
Other
0 stars 0 forks source link

Better method to obtain solar data? #2

Open taylormcd opened 6 years ago

taylormcd commented 6 years ago

Currently, this module relies on precomputed SMARTS data stored in specifically formatted csv files. Since SMARTS can only be used on Windows machines this significantly decreases the usefulness of this repository as a standalone package. My questions are these:

  1. Should we have a standard precomputed set of SMARTS data that accompanies this module?
  2. Where would we store this data?
  3. What format should we store this data in?

Or should we just assume that the user can use whatever method they can to obtain the necessary solar information, and that this package just gets them from that data to the solar capture on an aircraft? This might be the more appropriate decision, and would allow for this package to remain more general.

juddmehr commented 6 years ago

I have a proposed undergrad project to address this very issue. My hope is to bypass the need for SMARTS all together by creating a model based on precomputed data.

At the moment, it is assumed that users have their own datasets.

taylormcd commented 6 years ago

There might be other programs which compute the same data that SMARTS does as well. I remember at least one other source of data being used in papers which is not SMARTS. Perhaps that one might work as a SMARTS replacement.

taylormcd commented 6 years ago

In my opinion speed isn't an issue. The station keeping requirement restricts the latitude and longitude variation significantly so the only parameter that could change significantly when computing solar data is altitude. Even that difference probably has a minimal effect on the available solar flux. Therefore, I can't think of a case when solar data cannot be precomputed.

juddmehr commented 6 years ago

For a single aircraft at a single station, I agree, but when you're looking at trade studies for locations, seasons, etc. then it becomes important to have access to more data than just a single spacial point. This package may also be extended to applications outside High Altitude regimes, so if we're going to extend the functionality, we may as well go the whole nine yards.
That being said, I still can't think of a reason why any solar data couldn't be precomputed, even if there wasn't a station-keeping constraint, so I'm not particularly worried about speed either.