Setting value to variables using PyFMI

[lymereJ]

I built an FMU of the _fmu-export-variable.idf example file and I am using PyFMI to run the simulation. I've noticed that the value of yShadeFMU is always 0. I've tried to modify it using .set() (see here) but yShadeFMU stays at 0. I've tried doing the same using a FMU shipped with PyFMI, and there it worked. I'm concerned that there might something wrong with my FMU. I've tried on different operating system and get the same result. Am I missing something?

[tsnouidui]

Which version of E+ are you using? Could you try the same example using the _fmu-export-actautor.idf instead. Recent versions of E+ are not supporting erl-variables with the EnergyPlusFMU well. A similar issue has been reported in #16.

[tsnouidui]

@lymereJ any feedback on this? Otherwise, I will close the issue.

[lymereJ]

@tsnouidui, I have not tried your suggestion yet. I'll try this week and get back to you.

[lymereJ]

@tsnouidui, I'm using EnergyPlus 9.0 but I've also tried using version 8.9. I got the same results when trying the same example using the _fmu_export_actuator.idf.

[Denzo77]

@lymereJ, I had to pass yShade into the input parameter of model.simulate() to get it to work. I've uploaded an example here.

[lymereJ]

@Denzo77, thanks for getting back to me! I actually did try using input as an argument in model.simulate() using a slightly different approach than what you proposed in your gist (see below) and ... It was actually working 🙄. I was under the impression that it wasn't working because res['yShade'] is always return an array of 0s, I got so fixated on it that I didn't notice that it actually did have an impact on res['TRoo']. Anyway, thanks for bringing this topic back up and providing an answer!

For what it's worth, here's the approach I was following using the input argument:

def f(time):
    return 6

input_object = ('yShade', f)

res = model.simulate(start_time=0, final_time=60*60*24*3, input=input_object, options=opts)

@tsnouidui is it expected behavior to return an array of 0s?

[JChung-rpi]

Hi @lymereJ , I am wondering if you have solved this issue. I have been struggling with the same one for the past week. I have attached the figure I have generated using @Denzo77 's codes. I am using EnergyPlus v.9.4 which is the same version as the '_fmu-export-actuator.idf' file.

[lymereJ]

@JChung-rpi - Wow, that's a blast from the past!

I think that I did find a solution but I don't have that particular example anymore, I think., I've moved on to using the EnergyPlus Python API for newer projects. That being said, in a script that I still have on my machine (which is a bit more complex than the example) and also relies on PyFMI, it looks like I ended up using do_step() instead of simulate(), I think that was part of the solution since I see set() being used in the script.