(Question only)
Is there some idea for the Edition 2, how to model the "relative change of a measurand in some defined time interval" (gradient of a measurand) and also the status message, if some gradient-limit is exceeded ? For a lot of measurands the gradient is more important, than the absolute value, especially in monitoring systems.
Proposal
Discussion
Created
Status
09 May 07
In Force (green)
10 Apr 07
Ballot Period
10 Apr 07
Ballot Period
It is understood that these changes happen slowly and are used not for protection (like df/dt) but for monitoring. Therefore, this issue may be solved by adding a new calculation method ("trend")in the statistical nodes.