Opened 4 years ago
Last modified 3 years ago
#6264 assigned defect
An assert() is not correctly handled by OM: wrong violated equation
| Reported by: | Owned by: | Andreas Heuermann | |
|---|---|---|---|
| Priority: | high | Milestone: | |
| Component: | Backend | Version: | 1.16.0 |
| Keywords: | Cc: | Karim Adbdelhak |
Description
I more carrefully check results from simulation of models of the new library PowerSysPro developped by EDF and available here: https://bitbucket.org/simulage/powersyspro/wiki/Home
Important note: use the version 1.4 for OM (don't use the last version 1.5 due to discussion in ticket #6261).
The issue is related to the model OneSouceOneLineOneSource (package Examples).
The assert() condition is not correctly reported by OM as the log information is:
The following assertion has been violated during initialization at time 0.000000
src1.I <= 50.0
Current flowing the line is exceeding the maximum admissible current for PowerSysPro.Examples.OneSouceOneLineOneSource.line
The violated equation src1.I <= 50.0 is wrong and should be replaced by line.IA <= 50.0
This assert() is correctly handled by Dymola.
Change History (12)
comment:1 by , 4 years ago
| Status: | new → assigned |
|---|
comment:2 by , 4 years ago
follow-up: 7 comment:4 by , 4 years ago
The latest test is reported here: https://libraries.openmodelica.org/branches/master/PowerSysPro/PowerSysPro.html. Note that since mid-December, the new frontend is used by default for this test (master branch) and the newInst branch is obsolete and no longer run.
Apparently we are testing the master/HEAD version of your library in here, which is already using MSL 4.0.0. That is fine for us and should be independent of this specific issue. If you are interested, we can test both the latest stable release (1.14.0) and the development version, in that case please provide us the two URLs of the corresponding GIT repositories.
follow-up: 6 comment:5 by , 4 years ago
| Cc: | added |
|---|---|
| Component: | *unknown* → Backend |
| Owner: | changed from to |
Coming to your specific issue, I can confirm that we still get that with the 2.0.0 version of the library, see report.
I understand src1.I and line.IA are alias variables, i.e. they are identical. Of course from a documentation point of view it would be nicer to refer to the original variable, is that what you mean?
comment:6 by , 4 years ago
Replying to casella:
Coming to your specific issue, I can confirm that we still get that with the 2.0.0 version of the library, see report.
I understand
src1.Iandline.IAare alias variables, i.e. they are identical. Of course from a documentation point of view it would be nicer to refer to the original variable, is that what you mean?
Exactly, src1.I and line.IA are alias variables, and for the user it is a trouble as the assert() is done/tested in the line, not in the source. Should be nicer to refer to the original variable.
follow-up: 8 comment:7 by , 4 years ago
Replying to casella:
The latest test is reported here: https://libraries.openmodelica.org/branches/master/PowerSysPro/PowerSysPro.html. Note that since mid-December, the new frontend is used by default for this test (master branch) and the newInst branch is obsolete and no longer run.
Apparently we are testing the master/HEAD version of your library in here, which is already using MSL 4.0.0. That is fine for us and should be independent of this specific issue. If you are interested, we can test both the latest stable release (1.14.0) and the development version, in that case please provide us the two URLs of the corresponding GIT repositories.
There is no difference between PowerSysPro 2.0.0 and 1.4 except that with the latest 2.0.0 MSL 4.0.0 is used.
MSL version is independant from this issue. No need to test the previous 1.4 release.
follow-up: 9 comment:8 by , 4 years ago
Replying to jean-philippe.tavella@…:
There is no difference between PowerSysPro 2.0.0 and 1.4 except that with the latest 2.0.0 MSL 4.0.0 is used.
MSL version is independant from this issue. No need to test the previous 1.4 release.
Sorry, my question was independend from this specific issue. If you think that some people will still use 1.4 for some time, we should test that and make sure it works. If you have already deprecated it and have moved to 2.0.0 for good, then we can forget about it.
MSL 3.2.3 will be maintained and supported for a while. As a rule, I would keep on testing the latest MSL 3.2.3 version of each library we support for a few more years. Particularly before we have proper version management and conversion scripts in place, which is expected in version 1.18.0.
What do you think?
comment:9 by , 4 years ago
Replying to casella:
Sorry, my question was independend from this specific issue. If you think that some people will still use 1.4 for some time, we should test that and make sure it works. If you have already deprecated it and have moved to 2.0.0 for good, then we can forget about it.
Hum... my philosophy is to systematically use the latest stable versions (for tools, libraries and so on). I push my colleagues at EDF in this direction, with some disadvantages I know...
PowerSysPro 1.4 is to be seen as deprecated.
comment:11 by , 4 years ago
| Milestone: | 1.17.0 → 1.18.0 |
|---|
Retargeted to 1.18.0 because of 1.17.0 timed release.
New version PowerSysPro 2.0.0 is loaded in bitbucket.
This version is referring to latest Modelica version 4.0.0.