BookmarkSubscribe
Choose Language Hide Translation Bar
lily1
New Contributor

Finding an equation of a dynamic system

Hello Community,

 

 

I have 4 set of samples ( 2 input 1 output ) and I am trying to have the general equation representing the behavior of the dynamic system;  the equation that links all the 4 samples. How am i suppose to proceed ? 

 

0 Kudos
5 REPLIES 5
ian_jmp
Staff

Re: Finding an equation of a dynamic system

Your use of the phrases 'general equation' and 'dynamic system' might make one think that this is not a use case for JMP. So, what data do you actually have available, please? Folks will need to know this to offer any useful suggestions. 

0 Kudos
lily1
New Contributor

Re: Finding an equation of a dynamic system

Hello, 

 

all I have is the temperature input and the temperature output of the system and time

0 Kudos
gzmorgan0
Super User

Re: Finding an equation of a dynamic system

@lily1, a  few questions:

  1.  Is this sensor data, or read out data , that is why you call it dynamic?
  2. Do these 4 sample "sets" represent different time periods, tools or conditions? How much data in each set?
  3. Do you expect a common model ( pattern) for all 4 sets with maybe an offset or the same model?

If you can provide more detaisl, then the blog responders can help.

 

0 Kudos
lily1
New Contributor

Re: Finding an equation of a dynamic system

Hi @gzmorgan0

 

I will try to be more explicite. What I have is a Fluent model, in order for it to run I must provide Time and temperature of the heaters in his turn it provides the temperature of the water which are both time dependant. I have tried 4 different  heaters temperature  input and I have got 4 different outputs. Now without using the Fluent  simulation I am trying to find common model between the 4 samples. noting that the time periods are all the same and i have 8000 element in the input and output. Can JMP really help in this case ? 

 

Thank you Folks 

0 Kudos
gzmorgan0
Super User

Re: Finding an equation of a dynamic system

I still do not completely understand your question, or your problem. From your last post, it sounds like:

  • Time and temperature was entered into your fluent system, where time and temperature was constant over all 8000 elements.
  • There were 4 trials, each trial entered the same input time, only the temperature was varied.
  • Does that mean the data consists of 4 data points and you are trying to fit a model? Or are there multiple measurements of the output?

Note, with 4 points:

  • if the temperature deviations were small even a simple linear equation might not be discenible
  • the higest order polynomial would be 3.
  • if a model is found, it coud only be used to predict output with the lowets and hifest of your 4 samples and when all elements had the same temperature and the time used with your trials.

 

I have seen JMP used to simulate a multiple nozzle layout, using a characterized (modeled) flow rate by input. Then the system was optimized to set up custom flowrates to achieve its goal (across wafer uniformity). The engineers used JMP and JSL to create a simulator and perform a finite element analysis. However, the number of nozzles was closer to 100, not 8000, and it took JSL and engineering first principles and data.    

 

 

Sorry, it still is not clear to me.

   

0 Kudos