Math Is Fun Forum

  Discussion about math, puzzles, games and fun.   Useful symbols: ÷ × ½ √ ∞ ≠ ≤ ≥ ≈ ⇒ ± ∈ Δ θ ∴ ∑ ∫ • π ƒ -¹ ² ³ °

You are not logged in.

#1 2007-11-20 03:59:49

Const
Member
Registered: 2007-11-20
Posts: 2

Find relation/function/correlation

Hi,

I'm looking for a method to find a correlation/relation between test data.

I'm simulating a (compressed) air dyer.
I've got a mass load of calculations. Each calculation gives several results.
The first and most important result is the dew point, which represents the humidity of the air.
For each calculation there are also 4 other results: 4 temperatures.

Test nr.     dew point    T1     T2     T3     T4
1                 -25          10     20     30     40
2                 -35          44     88     99     66   
…
4000            -34          12     54     49     5

I’m looking for a method to (automatically) find the following function PDP = f(T1,T2,T3,T4), or any relation between the four temperature and the dew point.
When the relation is known, then I could predict the dew point when measuring those 4 temperatures.
My simmulation tool is written in Matlab. I would like to write a script that would find met the relation between the temperatures and the dew point.

Thanks for reading up,
Const

Last edited by Const (2007-11-20 04:02:48)

Offline

#2 2007-11-20 04:47:11

NullRoot
Member
Registered: 2007-11-19
Posts: 162

Re: Find relation/function/correlation

This might be far too simplistic, but have you tried seeing if the dew point can be the outcome of 4-variable simultaneous equations?

So for instance, with your example:
Test nr.     dew point    T1     T2     T3     T4
1                 -25          10     20     30     40
2                 -35          44     88     99     66   
4000            -34          12     54     49     5

Could you rewrite these as:
-25 = 10w + 20x + 30y + 40z
-35 = 44w + 88x + 99y + 66z
-34 = 12w + 54x + 49y + 5z

for instance? by solving what your variables (w, x, y, z) are you might find a relationship such that:
D = T1w + T2x + T3y + T4z
Where D is your dew point and T1, T2, T3, T4 are your observed/predicted temperatures?

As I say, it's probably grossly over-simplified, but it's probably worth a try if you haven't? At worst you'd be wasting very little of your time.


Trillian: Five to one against and falling. Four to one against and falling… Three to one, two, one. Probability factor of one to one. We have normality. I repeat, we have normality. Anything you still can’t cope with is therefore your own problem.

Offline

#3 2007-11-20 04:54:30

NullRoot
Member
Registered: 2007-11-19
Posts: 162

Re: Find relation/function/correlation

Ah. Hindsight. Always 20/20, aren't you, you little scamp!

I should have asked... What do the four temperatures have to do with the dew point? Or rather, what do they represent in this scenario?


Trillian: Five to one against and falling. Four to one against and falling… Three to one, two, one. Probability factor of one to one. We have normality. I repeat, we have normality. Anything you still can’t cope with is therefore your own problem.

Offline

#4 2007-11-20 19:28:18

John E. Franklin
Member
Registered: 2005-08-29
Posts: 3,588

Re: Find relation/function/correlation

Since you have 4000 results, I would sort them by dew point and spend many hours
gazing at all of the numbers, hoping that your brain will find some connection.
You knowing more about how this all works, might find a better way to examine the data,
such as noting that  T1 and T3 have a close connection or something...


igloo myrtilles fourmis

Offline

#5 2007-11-20 20:23:33

Const
Member
Registered: 2007-11-20
Posts: 2

Re: Find relation/function/correlation

I should have asked... What do the four temperatures have to do with the dew point? Or rather, what do they represent in this scenario?

The dryer works with a desiccant rotor. This rotor has a process side, where the wet air is going through to dry, and a regeneration side where hot air is blown through to dry the rotor.
The four temperatures are the temperatures of the air entering and leaving the process and regeneration side.

This might be far too simplistic, but have you tried seeing if the dew point can be the outcome of 4-variable simultaneous equations?

So for instance, with your example:
Test nr.     dew point    T1     T2     T3     T4
1                 -25          10     20     30     40
2                 -35          44     88     99     66   
4000            -34          12     54     49     5

Could you rewrite these as:
-25 = 10w + 20x + 30y + 40z
-35 = 44w + 88x + 99y + 66z
-34 = 12w + 54x + 49y + 5z

for instance? by solving what your variables (w, x, y, z) are you might find a relationship such that:
D = T1w + T2x + T3y + T4z
Where D is your dew point and T1, T2, T3, T4 are your observed/predicted temperatures?

As I say, it's probably grossly over-simplified, but it's probably worth a try if you haven't? At worst you'd be wasting very little of your time.

D = T1w + T2x + T3y + T4z could be possible but not likely.
There are to many possibilities like: D = T1w² + T2x³ + (T3y - T4z)²....

Therefore I was looking for some kind of program, script or method to automate it.

Thank you for the replys.

Offline

#6 2007-11-20 21:20:02

NullRoot
Member
Registered: 2007-11-19
Posts: 162

Re: Find relation/function/correlation

Const wrote:

The dryer works with a desiccant rotor. This rotor has a process side, where the wet air is going through to dry, and a regeneration side where hot air is blown through to dry the rotor.
The four temperatures are the temperatures of the air entering and leaving the process and regeneration side.

D = T1w + T2x + T3y + T4z could be possible but not likely.
There are to many possibilities like: D = T1w² + T2x³ + (T3y - T4z)²....

Therefore I was looking for some kind of program, script or method to automate it.

Thank you for the replys.

Hmm. If thats the case, then the form is actually probably something like D = f1(Ep,Lp) ? f2(Er,Lr)
(I renamed your variables to Entering or Leaving with a subscript of Process or Regeneration)
I would suggest there's probably a relation to the air that goes in and out of the P and R sides, but I couldn't make a guess at what operation(s) "?" should represent or what f1 and f2 are. As John says, you probably just have to stare at it and try different things that fit the context.

As far as my experience has been (which is by no means exhaustive), writing an program for this type of thing is hard because you pretty much need to tell the program what sort of things to look for in the first place.

Sorry I can't be of more help.


Trillian: Five to one against and falling. Four to one against and falling… Three to one, two, one. Probability factor of one to one. We have normality. I repeat, we have normality. Anything you still can’t cope with is therefore your own problem.

Offline

Board footer

Powered by FluxBB