Displaying attributes in PI Vision created by formulas in PI AF is too slow.

I am using PI AF to get an attribute value that I will display in PI Vision. The attribute is obtained from many formulas and table lookups. I am working on a heat exchanger "U" coefficient and it is calculated correctly, but I am having some problems when I want to show a trend for the last 2 years. I included attributes from daily TagAvg to reduce the amount of calculations, but this is not enough.

Parents
  • There could be a variety of reasons for things being slow. I see you used the tag "PI AF Analysis" but you also mention Formulas and Table Lookups, as well as a TagAvg.

     

    Depending upon the nature of a data request, the Formula Data Reference could be slow. And while many Table Lookups is usually fast enough, a poorly written one could also be slow.

     

    If performance is a concern, I would suggest migrating calculations to one or more AF Analytics and to persist the calculation(s) back to a PI Point. This is especially true for anything used for aggregation, e.g. tag averaging. Again, this may involve having staggered calculations to focus on and persist individual parameters that will be fed into the heat exchanger "U" cofficient calculation.

Reply
  • There could be a variety of reasons for things being slow. I see you used the tag "PI AF Analysis" but you also mention Formulas and Table Lookups, as well as a TagAvg.

     

    Depending upon the nature of a data request, the Formula Data Reference could be slow. And while many Table Lookups is usually fast enough, a poorly written one could also be slow.

     

    If performance is a concern, I would suggest migrating calculations to one or more AF Analytics and to persist the calculation(s) back to a PI Point. This is especially true for anything used for aggregation, e.g. tag averaging. Again, this may involve having staggered calculations to focus on and persist individual parameters that will be fed into the heat exchanger "U" cofficient calculation.

Children
No Data