Sign in to follow this  
Followers 0
Patriot

Factory Talk Historian SE Performance Equations

3 posts in this topic

I am looking to create a performance equation that is capable of determining a percentage of bad products during the last 15 minutes. I can create a performance equation that can divide the number of bad parts by the total number of parts, but I am not sure how to make this calculation only look at the data from the past 15 minutes. In other words, I need to update the number of bad products every minute, and the number of total parts every minute in Historian. I also want to update the rolling 15 minute calculation every minute. I want to look at the last 15 minutes of data every minute. I see this working in a way where if there were 10 bad parts, and then 15 minutes later, there was 30 bad parts, I would need to calculate that at the present time, there has been 30 - 10 = 20 bad parts over the last 15 minutes. This 15 minute window would be rolling every minute. For example, at 1:15 PM, I would be looking at the number of bad parts and total parts processed between 1:00 PM and 1:15 PM. At 1:16 PM, I would need to be looking at the range of 1:01 PM to 1:16 PM. I am not sure how something like this could be configured via a performance equation.

Share this post


Link to post
Share on other sites
Patriot, I have recently done something similar with parts run in the last hour as well as the last 15 minutes. I have not incorporated reject or bad part totals as of yet, as we don't have a means of pulling out reject tubes, unless the machine faults out. What I did, was in the PLC, I counted parts for a minute and then pushed that into a Fifo stack. My Fifo stack is pretty big as I'm taking the Data for the last hour. I then used Factorytalk, as I don't have Historian, to do the math. For the parts I ran in the last hour, I just add each "entry" in the Fifo stack together. This number always varies unless the machine has been running continually for more than an hour. As for the last 15 minute run, I just use the most recent 15 entries in the Fifo stack to get my figures. As for your equation on determining bad part percentage, I would use 2 Fifo stacks, one for total parts and one for bad parts. I would then divide my "last 15 minutes worth of bad parts" total by my "last 15 minutes of good parts" total. I hope I've read what you want to do right and that this helps. Phil Edited by Phil Izak

Share this post


Link to post
Share on other sites
Phil, Thanks for the suggestion. I gave some thought to doing this in the PLC, and after your suggestion, I might end up doing this in the PLC similar to what you described. Once I have the data for the past 15 minutes in a tag in the PLC, it is easy to get it back to FT Historian. If I do the calc in FT Historian through a complex performance equation, I then need to write this data back to the PLC. If I just do this all in the PLC, it takes that extra step out. I will see if the customer is ok with this. The PLC was being used basically as a "data concentrator" for older PLC-5s, etc. to get the plant data in a ControlLogix platform. Thanks again.

Share this post


Link to post
Share on other sites

Create an account or sign in to comment

You need to be a member in order to leave a comment

Create an account

Sign up for a new account in our community. It's easy!


Register a new account

Sign in

Already have an account? Sign in here.


Sign In Now
Sign in to follow this  
Followers 0