Your question for the problem formulation is not clear enough?

I understood that you have an equivalent circuit model for a single cell (????)and you have measured impedance spectrum (?????) and you try to use the curve-fitting to find the parameters of the equivalent circuit that best fit the measured values. You have two tracks.

First:- if you have closed formula that ties the circuit element with the spectrum you can find the circuit elements for each spectrum and the circuit element is the average value for the hole spectrum.

second:- you can solve the problem as a whole problem by constructing a composite objective function with different weights that best fit the element values to the calculated spectrum.

Good Luck.

Prof. S. EL-Rabaie

Prof. S. EL-Rabaie is right that are your options. I want just to expand a little more. In Matlab, you can call the function fminsearch(f,x0) with your specific objective function and a initial guess. with your data it will find the best fit for your data and return a vector of parameters X. There is a third option but is more complex and suffer from horizon limitations. You can use neural networks of GP to learn a model over your data.

Thank you Prof. for the advise.