Problem with GlobalSearch - Algorithm evaluates only initial point

Hi,

I'm trying to minimize the distance between two matrices using Globalsearch (Distance: A*Omega*A'-Sigma). Unfortunately, the algorithm doesn't seem to proceed, instead it evaluates the objective function at the same (initial) point at every run. Does anyone have an idea why that could be the case?

My optimization looks like this:

 for i=1:numiter
    disp(['Iteration No.',num2str(i),' in progress'])
    SV;
problem = createOptimProblem('fmincon',... 
                            'objective',@(SV) objectivefun(SV,S,N,cov_r1,cov_r2,cov_r3,cov_r4),...
                            'x0', SV,...
                            'options', optimset('MaxFunEval',inf,'MaxIter',Inf, 'LargeScale', 'off'));
gs = GlobalSearch('NumTrialPoints', 3000, 'Display', 'iter');
[estimator, fval]=run(gs,problem);
    if fval<1e-002        
        disp('Converged!')
       break
    end
    SV=estimator;
end

and my objective function is

 function x = objectivefun(SV, S, N, cov_r1, cov_r2, cov_r3, cov_r4)
As = eye(N);
As(2:3,1) = SV(1:2,1); As(1,2) = SV(3,1); As(3,2) = SV(4,1); As(1:2,3) = SV(5:6,1);
As=kron(eye(S),As);
Sigmas = diag(SV(N*N-(N-1):end,1));              
Omega = blkdiag(cov_r1, cov_r2, cov_r3, cov_r4);
x = sumabs(As*Omega*As'-Sigmas);      
end

S and N are scalars, 4 and 3, respectively. cov_r1...cov_r4 are 3x3 Matrices. So that Omega, Sigma and A are all 12x12. It is supposed to optimize only over certain elements in A and the diagonal elements of Sigma - that is why I allocated the starting values to values in A and Sigma.

If someone has any idea what could be wrong, I'd greatly appreciate it.

Thanks a lot in advance. Regards!