As for your expressions, I personally would prefer PredictiveTot_Of_The_Current_Shift_1 as it clearly illustrates your intent, which is you have a partial shift total and you want to scale it out for the full shift.
From Live Library Help for PI Server 2016 R2, there is a TagTot function for tags and a TagTot function for attributes. Both expect the input item to be expressed in units per day. If your 'Debit' is not a daily rate, you will need to apply a conversion factor to get the right value.
For PI Server 2017, there is no change with TagTot for tags, but there is a new parameter for TagTot for attributes, where you may specify the desired output UOM. If it's compatible with the inputs, the conversion is performed for you.
TagTot returns a number without a timestamp. To save the results back to PI, the Asset Analytics Output Time Stamp applies (from the Advanced... button). That is you the assigned output time stamp may be one of 3 things:
- Trigger Time
- Execution Time
- Relative to Trigger Time
If you choose Relative to Trigger Time, as you type in the box and enter *+ the dialog automatically senses this will be a future timestamp (since you are adding to Now), and presents you with this text:
For possible language translators, that would read:
Mapped output attributes require a future PI point to save output history. For newly mapped output attributes that save output history, PI Analysis Service will create future PI points.
thank you very much for replying to my discussion and for your explanations.
Is your total over time a linear function?
no it isn't.