David Garcia
unread,Jul 14, 2010, 5:50:06 PM7/14/10Sign in to reply to author
Sign in to forward
You do not have permission to delete messages in this group
Either email addresses are anonymous for this group or you need the view member email addresses permission to view the original message
to DXC10
[Taken from a email bug report. The patch mentioned in the response
will be posted shortly.]
> Wilson, Michael A wrote: Hello,
>
> I was looking through the "DXC10_IndustrialTrackDescription.pdf"
> document posted on the DX competition website, and I noticed a
> discrepancy between the Oracle and "Table 10. Diagnostic Problem II
> Recovery by Failure Modes".
>
> With ADAPT in configuration 10, I sent the Oracle a query for the
> proper response to a FAN415.UnderSpeed fault. According to Table 10
> in the aforementioned document, there is no recovery action
> available. However, the Oracle responded with a recovery command
> suggestion. The following is an excerpt from the result scenario that
> I ran to test this (the first line is the query sent to the Oracle by
> the DA, and the second line is the Oracle's response):
>
> ... recoveryInfo @1278544295908 <DA> faults = { FAN415 = UnderSpeed
> }, commands = { }, cost = -1, notes = ''; recoveryInfo
> @1278544296111 <OR> faults = { FAN415 = UnderSpeed }, commands = {
> EY275_CL = true }, cost = 1, notes = ''; ...
>
> Were there possibly changes made to the Oracle that are not reflected
> in the Industrial Track description document? Or is Table 10
> referring to some specific configuration of ADAPT that does not match
> my test configuration?
>
> Thank you for your time.
>
> Best regards, Mike Wilson
>
>
Mike,
Yes, it looks like this is a bug in the oracle. It looks like the
constraints in the underlying model for the fan are reversed:
Overspeed was not requiring a recovery, and Underspeed was requiring a
recovery. It should be the other way around.
I created a patch for this bug. To install it, copy the two files
attached to the email into your DXC_HOME\bin\models. You'll overwrite
the two files which already exist in that dir. The unit-level tests I
created show that this should fix the problem (output below). For the
DXC team, we should decide how to handle the bugfix. Probably
supplying these new model files as a patch is good enough for now. If
we do make another full release, we would incorporate this patch. I
checked it into the DXC svn.
For fault set : { FAN415 = OverSpeed }
{ EY275_CL = true } computationTime = 0.765
For fault set : { FAN415 = FailedOff }
{ EY275_CL = true } computationTime = 0.766
For fault set : { FAN415 = UnderSpeed }
{ systemAction = 'NOP' } computationTime = 0.531
Adam