[SOLVED] Fitting a cole-cole model to data

[lucy123]

Hi all,

I have obtained the dielectric properties of biological tissue and I am trying to fit a cole-cole model to the dielectric const.

function diff = function_110502(x,freq)
diff = A+((B-A)./(1+(2*pi*freq*C).^(2*(1-D))))

X0=[set of initial conditions]';
[x]=lsqcurvefit(@function_110502,X0,freq,Y)

I have tried using lsqcurvefit but it giving me back the initial values. I have also tried to use the cftool directly using the non-linear least square but still unsuccessful.

Can anyone help pls?

 

[Prof78]

There is an article on the website The Designer's Guide Community - A Resource for Analog, RF, and Mixed-Signal Circuit Designers by Ken Kundert entitled 'Modeling dielectric absorption in capacitors' that may be of use. It includes a number of references. Guessing from the nature of your enquiry it would seem that you may have access to an institutional library. There is an annual series of books entitled Progress in Dielectrics which may may contain relevant information. Reference [6] in Kundert for example is to the Cole-Cole theory appearing in Progress in Dielectrics.