Hi all. I've implemented a model that needs 13 inputs parameters to work. Now, I am trying to change these 13 input parameters to get simulation results of this model close to experimental results. Actually, I need to play with these 13 parameters so as the simulation results and experimental results fit (a really hard work by hand). I don't know what should I do? I wonder if somebody can give me some advice. Is there any calibration or modification function and/or software that I can use for this purpose? best,
Would Minerr help? In this example Z is a function that is the output you are after. Z_data is a set of data. Mathcad optimizes the variables to find the minimum error.
I'm sorry, but aren't your two weightings the same?
Some simple algebra manipulation of "Normalized weighting' results in "Natural weighting."
What did I miss?
You are right, one good approach (especially with emperical data) is Minerr!
Except that nothing actually equals zero since it's (in my case) a fit of data. We're just trying to get close to zero. I certainly get different answers with the two methods.
Might also try
Fn_data/Data = 1
That gives the same result as the 'normalized.' And that makes sense because they are the same equation.
Maybe, are you looking for some fitting routine(s)? Attached is P3.1 sheet for your reference.