Missing nominal attribute in integer tearing variables nlsData C structure

[casella]

Please have a look at the result of this test

​https://test.openmodelica.org/libraries/ThermoPower_Experimental/files/ThermoPower.Examples.HRB.Simulators.ClosedLoopSimulator.err

two integer variables are selected as tearing variables: $cse129.phase and $cse129.region

Wrong C code is generated as a result:

ThermoPower.Examples.HRB.Simulators.ClosedLoopSimulator_02nls.c:6513:54: error: no member named 'nominal' in 'struct INTEGER_ATTRIBUTE'
  nlsData->nominal[i] = $P$ATTRIBUTE$P$cse129$Pphase.nominal;
                        ~~~~~~~~~~~~~~~~~~~~~~~~~~~~ ^
ThermoPower.Examples.HRB.Simulators.ClosedLoopSimulator_02nls.c:6517:55: error: no member named 'nominal' in 'struct INTEGER_ATTRIBUTE'
  nlsData->nominal[i] = $P$ATTRIBUTE$P$cse129$Pregion.nominal;
                        ~~~~~~~~~~~~~~~~~~~~~~~~~~~~~ ^
2 errors generated.

This error popped up between Jul 6th and Jul 26th.

[lochel]

What I can say just from the error message: It seems to be an issue with mixed systems – that means systems with both discrete and continuous iteration variables. That is currently not supported. The only good thing about it is, that it is already on my list ;-)
Anyway I can have a look at it tomorrow.

[sjoelund.se]

Moved to new milestone 1.9.4

[casella]

Somehow this issue was resolved between Sep 10 and Sep 14.

I am closing the ticket.

[lochel]

I think this was somehow effected by separating the transformation of simulation and initialization. Maybe there is now some simplification applied to the system that omit the error.

[casella]

So, mixed systems are still not supported, basically?

[lochel]

Replying to casella:

So, mixed systems are still not supported, basically?

Yes, mixed systems are not handled in general (see also #3440). However, it is possible to solve those systems in some cases at least, due to a hack in the nonlinear solvers.

[sjoelund.se]

Milestone renamed

[sjoelund.se]

Milestone renamed