Time-Weighted vs Event-Weighted Average
How PI AF computes averages — event-weighted vs time-weighted, continuous vs discreet.
How PI AF computes averages — event-weighted vs time-weighted, continuous vs discreet.
The following tags are used for this example: ExampleTag1 and ExampleTag2. The table below shows the values in the archive for the two tags over a one-minute interval. ExampleTag1 has an archive value every 10 seconds, whereas ExampleTag2 has the same values in the archive but stored at varying intervals.
When using the PI Tag data reference in PI AF to retrieve the average value for the same minute as the table above, the available summary types are:
Event-Weighted — weights each event equally and uses events at each boundary.
Event-Weighted Exclude Earliest Event — same as Event-Weighted, except the event at the start (earliest) time is not used.
Event-Weighted Exclude Most Recent Event — same as Event-Weighted, except the event at the end (most recent) time is not used.
Event-Weighted Include Both Ends — same as Event-Weighted.
Time-Weighted — weighs each event value by the length of time over which it applies.
Time-Weighted Continuous — time-weighs the values, using linear interpolation between values regardless of whether the attribute is configured as stepped.
Time-Weighted Discreet — time-weighs the values, using stepped interpolation between values regardless of whether the attribute is configured as stepped.
The results ignore any time effect, since they use only the event values and nothing else.
The result is calculated as the sum of the areas (A1 – A6) divided by 60 seconds.
Talk to our team about a tailored AF assessment.