Explore the FMU

Knowledge about the FMU is necessary to setup a probabilistic approach:

- What is the name of the FMU variables ?

- Which are the inputs, outputs, parameters ?

- Which are booleans, reals, integers ?

- What is their default start value ?

First, retrieve and load the FMU deviation.fmu.

import otfmi.example.utility

path_fmu = otfmi.example.utility.get_path_fmu("deviation")
model = otfmi.fmi.load_fmu(path_fmu)

Let require the FMU variables names:

list_name = otfmi.fmi.get_name_variable(model)
print(list_name)

['der(_D_outputAlias_y)', 'E', 'F', 'I', 'L', 'y']

This command shows all variables : inputs, parameters, outputs.

Let identify their causality in the model:

for name in list_name:
    causality = otfmi.fmi.get_causality_str(model, name)
    print("{}: {}".format(name, causality))

der(_D_outputAlias_y): LOCAL
E: INPUT
F: INPUT
I: INPUT
L: INPUT
y: OUTPUT

Yet the variables type is not known: real, integer, boolean, string?

Let check using PyFMI’s method:

for name in list_name:
    typ = model.get_variable_data_type(name)
    print("{}: {}".format(name, typ))

der(_D_outputAlias_y): 0
E: 0
F: 0
I: 0
L: 0
y: 0

The type 0 corresponds to Real (aka “float”) variables.

Let check the variables default start value in the FMU:

dict_start_value = otfmi.fmi.get_start_value(model)
print(dict_start_value)

{'E': 30000000.0, 'F': 30000.0, 'I': 400.0, 'L': 250.0}

Note

Function otfmi.fmi.get_start_value only returns the start value of variables with types Real, Integer or Boolean.

With this knowledge on the FMU variables, we can now simulate it (with non-default initialization values if required).