Opened 4 years ago
Last modified 3 years ago
#6019 new defect
previous(y) is chattering in a discrete-time model
| Reported by: | Owned by: | Lennart Ochel | |
|---|---|---|---|
| Priority: | critical | Milestone: | |
| Component: | Run-time | Version: | v1.16.0-dev |
| Keywords: | Cc: |
Description
Hi,
I built a small wrapping counter with time-discrete blocks.
When simulating in openmodelica, One of the available outputs is previous(y).
As you see in the attached screenshot, the variable behaves very strange, quite different from y itself.
Additionally I'm confused by the fact that variables i and y are calculated between sampling ticks. This is not how I understand how it was intended (https://ep.liu.se/ecp/076/001/ecp12076001.pdf)
Best regards,
Paul van der Hulst
Attachments (3)
Change History (7)
by , 4 years ago
| Attachment: | TDMclock2.mo added |
|---|
by , 4 years ago
| Attachment: | screenshot.png added |
|---|
comment:1 by , 4 years ago
| Component: | *unknown* → Run-time |
|---|---|
| Milestone: | Future → 1.16.0 |
| Owner: | changed from to |
| Priority: | normal → critical |
| Version: | v1.14.1 → v1.16.0-dev |
There is clearly some bug in the recording of previous() variables. If you look at the detail in the plot.png attachment, at some point it actually goes backwards in time. Nice picture, vaguely reminding a Mondrian picture, but completely meaningless.
@Karim, @AnHeuermann, would you mind giving a look?
by , 4 years ago
comment:3 by , 4 years ago
| Milestone: | 1.17.0 → 1.18.0 |
|---|
Retargeted to 1.18.0 because of 1.17.0 timed release.
Screenshot