ModelStructure with wrong dependencies for states and without dependencies for outputs

[rfranke@…]

Using the current nightly build 1.9.1+dev (r21649), the following model

within ;
model DIC "Double Integrator Continuous-time"
  parameter Real p = 1 "gain for input";
  parameter Real y1_start = 1 "start value for first state";
  parameter Real y2_start = 0 "start value for second state";
  input Real u(start = -2);
  output Real y1, y2;
initial equation
  y1 = y1_start;
  y2 = y2_start;
equation
  der(y1) = p*u;
  der(y2) = y1;
end DIC;

can be exported as FMU 2.0 with the omc commands:

    loadFile("DIC.mo");
    translateModelFMU(DIC, version="2.0");

The file modelDescription.xml contains the following ModelStructure:

  <ModelStructure>
    <Outputs>
      <Unknown index="1" dependenciesKind="fixed" />
      <Unknown index="2" dependenciesKind="fixed" />
    </Outputs>
    <Derivatives>
      <Unknown index="3" dependencies="1" dependenciesKind="fixed" />
      <Unknown index="4" dependencies="2" dependenciesKind="fixed" />
    </Derivatives>
    <InitialUnknowns>
    </InitialUnknowns>
  </ModelStructure>

The Outputs section lists dependeciesKind, but is lacking the dependencies (indices). The Derivatives section states wrong dependencies. The expected ModelStructure is:
\partial y / \partial x
\partial y / \partial u

\partial der(x) / \partial x
\partial der(x) / \partial u

  <ModelStructure>
    <Outputs>
      <Unknown index="1" dependencies="1" dependenciesKind="fixed" />
      <Unknown index="2" dependencies="2" dependenciesKind="fixed" />
    </Outputs>
    <Derivatives>
      <Unknown index="3" dependencies="5" dependenciesKind="fixed" />
      <Unknown index="4" dependencies="1" dependenciesKind="fixed" />
    </Derivatives>
    <InitialUnknowns>
    </InitialUnknowns>
  </ModelStructure>

[Willi Braun]

fixed in r22294.

[Rüdiger Franke]

Now the Derivatives section looks good for the given example! The Outputs section still is wrong though:

  <ModelStructure>
    <Outputs>
      <Unknown index="1" dependencies="5" dependenciesKind="dependent" />
      <Unknown index="2" dependencies="1" dependenciesKind="dependent" />
    </Outputs>
    <Derivatives>
      <Unknown index="3" dependencies="5" dependenciesKind="dependent" />
      <Unknown index="4" dependencies="1" dependenciesKind="dependent" />
    </Derivatives>
    <InitialUnknowns>
    </InitialUnknowns>
  </ModelStructure>

Note: the Outputs section looks good if additional variables are introduced for states, so that state variables differ from outputs.

[Willi Braun]

This would mean I need to ignore the output declaration in the model, is this desired?

Since I really haven't found any specification text that a variable can't be a state and a output
at the same time.

[Rüdiger Franke]

A variable can generally be state and output. See also Modelica.Blocks.Continuous.Integrator. I had discussed this in the FMI group and was told that FMI has intentionally be defined in a way supporting such models. Such outputs depend on themselves, i.e. output with index 1 depends on state with index 1 and output with index 2 depends on state with index 2.

[Willi Braun]

Doh, you are of course right the output section is wrong. Somehow I mixed there somewhere states and states derivatives. I'll check that.

Replying to rfranke:

Note: the Outputs section looks good if additional variables are introduced for states, so that state variables differ from outputs.

Now, I also get the idea of that comment.

[Adeel Asghar]

While we are fixing ModelStructure we should include InitialUnknowns & DiscreteStates.

I have added some support for InitialUnknowns in r22421. Also see #2765.

[Martin Sjölund]

This ticket was not closed for 1.9.1, which has now been released. It was batch modified for milestone 1.9.2 (but maybe an empty milestone was more appropriate; feel free to change it).

[Martin Sjölund]

Milestone changed to 1.9.3 since 1.9.2 was released.

[Martin Sjölund]

Moved to new milestone 1.9.4

[Martin Sjölund]

Milestone pushed to 1.9.5

[Martin Sjölund]

Milestone renamed

[Martin Sjölund]

Ticket retargeted after milestone closed

[Martin Sjölund]

Milestone moved to 1.12.0 due to 1.11.0 already being released.

[Willi Braun]

Fixed in ​OMCompiler/bba43625.