Re: using fsolve returns initial values

"Magnus" wrote in message <lu1j4m$jq1$1@newscl01ah.mathworks.com>...
> Dear all,
>
> I'm trying fsolve to solve a nonlinear system comprising of 4 equations.
> The function that I declared looks like this:
> %%%%%%%%%%
> function F = myfun( y )
> %y is the solution vector (xx zz zeta1 eta1)
>
> F=[y(4)-A2*sinh(2*k*(d+eta2i))/sinh(2*k*d)*cos(2*k*zeta2i)-B2-eta2i;
> y(1)-A1*cosh(k*(d+y(4)))/sinh(k*d)*sin(k*y(3))-y(3);
> y(2)-A1*sinh(k*(d+y(4)))/sinh(k*d)*cos(k*y(3))-y(4);
> y(3)-A2*cosh(2*k*(d+ea2i))/sinh(2*k*d)*sin(2*k*zeta2i)-zeta2i];
> %%%%%%%%
> All variables are declared global and there are only four unknown variables. Vector y is the solution vector and supposed to give the four values xx zz zeta1 and eta1.
>
> Now, I'm using the following code to call the function.
> The loop is used to make the calculation for 120 times.
> %%%%%%%%%%%%
> x00=[1;1;1;1]; % Make a starting guess at the solution
> for i=1:120
> global eta2i
> eta2i=eta(i);
> global zeta2i;
> zeta2i=vector(i)
>
> options = optimoptions('fsolve','Display','iter'); % Option to display output
> [y,fval] = fsolve(@TransformZetaAndEta,x00,options); % Call solver
> xx(i)=y(1);%y is vector (xx zz zeta1 eta1)
> zz(i)=y(2);
> zeta1(i)=y(3);
> eta1(i)=y(4);
> end
> %%%%%
> However, when I have a look at the solution, the vectors xx, zz, zeta1 and eta1 are all the initial values, i.e. 1. Changing the initial value-vector x00 doesn't make any difference. However, these values can't be valid or a true solution to the problem (I tried to make a calculation using these values which results in error messages).
>
> I would very appreciate any help or hint to this problem.
>
> Kind regards and thanks in advance
> Magnus

Why do you use fsolve to solve your System of equations ?
From equation 1 and 3, you directly get y(3) and y(4).
Inserting these values in equations 2 and 3, you get y(1) and y(2).

Best wishes
Torsten.