how to solve a deflection problem in pavement surface using genetic algorithms.

The question is like this,

I have the formula for the theortical deflection d(t) that depends on four parameters(u,v,x,y) that have a specific range ( let it be theoretical deflection)

along with these, I have six measured deflections d(m) at six different points.

So, I need a GA

where the function is the root square mean of the sum of error in each deflection.

i.e.,

Where the outcome should have a single value for every parameter so that the error should be minimized and have exact or nearby value to measured deflection.

[ADITYA ANUPAM]

Somebody Please try to solve this It will be really helpful

[darova]

Why there are no u, v, x and y in formula?

[ADITYA ANUPAM]

No No formula

[ADITYA ANUPAM]

No, Constant value need to be find out from theoritical deflection.

Answer 1

[Alan Weiss]

There are many ways to solve such model-fitting problems. See these examples:

https://www.mathworks.com/help/matlab/math/example-curve-fitting-via-optimization.html

https://www.mathworks.com/help/optim/ug/nonlinear-data-fitting-example.html

https://www.mathworks.com/help/gads/multistart-using-lsqcurvefit-or-lsqnonlin.html

You can always use ga instead of fminsearch or fminunc, though I am not sure why you would want to.

Alan Weiss

MATLAB mathematical toolbox documentation

[ADITYA ANUPAM]

Sir,

If these question were solved by using fminsearch I had used it earlier. It's my research work that's why i can't explain here. But this example is just a protype of what I want. So if you can help me to solve I will be highly obliged to you.