I need a mutation function with good convergence but working with bounded individuals bound

Hi all

I'm developing a code for finding fitting parameters of a polynomial.

When I used the gaussian mutation function it gave a good conversion as shown in the figure (bottom image) but the upper image shows the problem of unbound negative solution, I need the gaussian convergence but with bound individuals.

By using the mutationadaptfeasible the convergence was slow, Is there any options for controlling the mutationadaptfeasible convergence?

I'm using Matlab R2018a

[John D'Errico]

The thing is, I cannot figure out why you want to fit a polynomial using a genetic algorithm. Given that there are vastly bettter schemes to estimate the coefficients of a polynomial fit, why? Let me guess. A homework assignment, designed to make you work like hell to do a poor job on something you would not want to do in practice, in a million years?

[Ezzat]

So, what is your recommendations for polynomial fitting of degree 4 ??