Fitting an custom equasion where i only want to optimize specific parameters

I want to optimize the fit of my custom equasion. However I only want to change 3 spezific parameters, but the funktion has multiple parameters.

My funktion looks something like this:

1000000*sigk*h1*g1*(c3+(1-((H./100).^c4))/(c5*(1+(Dwir/100)))).*f1)

with sigk, h1, g1, Dwir = skalar

H is a vektor 1x365

f1 has previously been calculated using:

f1 = zeros(1,length(t));

for m = 1:5 %Schleife verschiedene Kelvinelemente

f1 = f1+n(1,m)*(1-exp(-a(1,m)*(t.^b_f1-tau^b_f1))); %Zeitfunktion, Vektor

end

I want to fit the parameters c3,c4 and c5

I have a Vektor "Data" with data from an experiment.

Can you explain to me how I can use the Curve Fitting app or if that is even possible? thank you.

Answer 1

[Star Strider]

You would need to supply pre-determined values for ‘sigk’, ‘h1’, ‘g1’, and ‘Dwir’ if you have them (you have already calculated ‘f1’), so substitute them into your function and then estimate ‘c3', ‘c4’and ‘c5’.

I assume ‘Data’ and ‘H’ are the same size, so you could then use your function to estimate the desired parameters.

The function is nonlilnear in the chosen parameters ‘c3', ‘c4’and ‘c5’, so you would need to use a nonnlinear parameter estimation function. There are several available in MATLAB.