Curve Fitting for Forced Damping

Hello,

I used the curve fitting app to fit this equation to some data:

F./(m*sqrt((w0^2-w.^2).^2 + (g^2)*w.^2))

The constants for the equation are F, m, g and w0. w is the independent variable.

My starting points for the data were F = 0.8491, g = 0.9340, m = 0.6787, and w0 = 125.

The method chosen is "NonlinearLeastSquares" and the algorithm chosen is "Levenberg-Marquardt".

However, the fit doesn't fit all that well. Here's a picture.

Any tips on how to improve this fit? The R squared value for this fit is around 0.73.

My start points for this fit we

Thanks!

Answer 1

[Alex Sha]

Where are your data?