Refine the results by GA to true values

I am currently using GA optimization toolbox.

My true values to generate my dataset is, say like [0.4 0.5 0.6],

after i run my simulation, it gives me [0.9, 3.5 0.2],

What to do in order to obtain the result near to my true values?

[Walter Roberson]

Could you show a bit of your code?

[Khoo]

This is my constrained function file:

    function [c,ceq]=nonlincon(z)

    c=z(3)*exp(z(2)*(1-z(1)))-1;
    ceq=[];

this is my main file

for iter=1:iterate   
z=[0.3 0.5 0.4];  
   options=gaoptimset('TolFun',1.0000e-     008,'Display','iter','UseParallel','always','InitialPopulation',z);
h=@mptest;
contraint=@nonlincon;
est=ga(h,3,[],[],[],[],[0 0 0],[1.0 10.0 1.0],contraint,options);
sumest=sumest+est;
SE=(est-z).^2;
sumSE=sumSE+SE;
fprintf('Iteration= %1.0i\n',iter);
end

I am doing the monte carlo simulation. Is that possible to find the estimated parameters near the initial point?