dpinney / omf

The Open Modeling Framework for smart grid cost-benefit analysis.
https://omf.coop
GNU General Public License v2.0
112 stars 60 forks source link

New blank feeder creation. [20h] #106

Closed dpinney closed 9 years ago

dpinney commented 11 years ago

TODO

THOUGHTS Currently we only allow Milsoft and GridlabD import for feeder creation. A blank feeder capability would be useful for tests and modeling efforts without Milsoft data.

dpinney commented 10 years ago

Deferred.

dpinney commented 9 years ago

Assigned.

mannanj commented 9 years ago

What is the bare-minimum for a new feeder? Going off our simplest feeder, the Simple Market System feeder, I reduced it one house further and am thinking of the following:

newfeeder

Generating the feeder takes the user directly to that screen (slightly zoomed off to the top right of the screen). It seems the zoom to fit button doesn't work, I don't think this is too major so seems it won't be taking the user directly to the view seen in the screenshot.

dpinney commented 9 years ago

Minimal feeder is probably just the pink dot (swing node) and the gray disconnected nodes.

Does zoom reset work better than zoom to fit? I agree that zooming for the new feeder is not critical to fix now.

mannanj commented 9 years ago

Zoom reset doesn't do it either. This is now working in gridlab multi & will be closed when its committed later tonight.

FUTURE TODO: Fix zoom so it shows something like the above view, and locks on the feeder.

dpinney commented 9 years ago

Could you take out the second transformer and meter circled in the below?

untitled

mannanj commented 9 years ago

Seems I can't, because taking it out crashes the gridlabmulti result. As soon as I put it back in, it works.

I thought it could be from leaving the transformer configuration in there, so I took it out. But that didn't resolve it. Not sure why this is happening.

dpinney commented 9 years ago

What is the error message?

mannanj commented 9 years ago

MODEL CRASHED GRIDLABMULTI unsupported operand type(s) for /: 'str' and 'int' /home/samsung/Documents/Projects-active/omf/omf/data/Model/admin/SimpleFeeder SimpleFeeder

CANCELED /home/samsung/Documents/Projects-active/omf/omf/data/Model/admin/SimpleFeeder/SimpleFeeder

MODEL CRASHED GRIDLABMULTI local variable 'feederOutput' referenced before assignment /home/samsung/Documents/Projects-active/omf/omf/data/Model/admin/SimpleFeeder

CANCELED /home/samsung/Documents/Projects-active/omf/omf/data/Model/admin/SimpleFeeder

dpinney commented 9 years ago

Check the stderr.txt in the model's subfolder. What does that say?

mannanj commented 9 years ago

Aha that has pinpointed the error to line 158.

It says: Traceback (most recent call last): File "/home/samsung/Documents/Projects-active/omf/omf/models/gridlabMulti.py", line 158, in runForeground cleanOut['allMeterVoltages']['StdDev'] = hdmAgg([float(i / 2) for i in rawOut['VoltageJiggle.csv']['std(voltage_12.mag)']], avg, level) TypeError: unsupported operand type(s) for /: 'str' and 'int'

Its because of the two types in and str. I will debug this now.

dpinney commented 9 years ago

Okay. If you get stuck let me know and I'll share my screen.

mannanj commented 9 years ago

Seems its actually because the std column in VoltageJiggle.csv is nan. Shall I debug further?

nan

mannanj commented 9 years ago

The file voltagejiggle.csv is generated in gridlabD, when I try a simple try and except pass it'll work but it throws off some of the outputs in the gridlabmulti model. That output looks like this: (the min/max/avg for this reduced feeder are wrong, they are all the same in the chart)

gridlabd1

Would you know if there is a minimum feeder size that gridlabD must have in order to run smoothly?

dpinney commented 9 years ago

The issue is with the GLD output cleaning code. Just leave the 2nd meter on the model.

mannanj commented 9 years ago

Okay that has been done and committed.

dpinney commented 9 years ago

I tested this and confirmed it works. Nice job.