Did you run residual analysis approach? in the simplest way, the best fit is one that have minimum residual (minimum root mean squares) relative to M curve.

I did not try this. Thank you very much for your suggestion, but I would like to convolute all the curves to obtain the bold one.

You just give these curves for the graph curve readers program that tabulate these graphs.

You will construct the residual function which is the difference between the values for all the points to each curve and the bold one.

The lowest residual is the more nearest to the bold one.

If you want to con-volute all the curves to obtain the bold-one that will be easier if you look carefully for the tabulation values of the curves.

Prof. S. EL-Rabaie

Dear Prof. S. El-Rabaie

Thnak you very much for your help

Best regards

Malgorzata

There is a lot of documentation for (interactive) curve fitting:

https://de.mathworks.com/help/curvefit/curve-fitting.html