Multiparametric optimisation problem using Genetic algorithm

I'm currently trying to optimise a velocity demand signal with the values to be optimised being 'POINT' based on the cost function 'mycostfunction'

I've checked the value of cost changes if i differ my 'POINT' numbers but when I try to use a GA code the system crashes and when I use ga toolbox the system seems to get stuck and not get past its first iteration I am really confused can someone please help me

if true
  % function cost = mycostfunction(POINT)
global POINT
global tstep
global veldem
global bspliney
global bsplinex
global response1
global controlpoints
global xycoords
global ycoords
% POINT=[48 48 48 48 48 48 48 48 48 28 28];

xycoords = [0 POINT(1); 36 POINT(2); 72 POINT(3); 108 POINT(4); 144 POINT(5); 180 POINT(6); 216 POINT(7); 252 POINT(8); 288 POINT(9); 324 POINT(10); 360 POINT(11)]; ycoords = [POINT(1); POINT(2);POINT(3);POINT(4);POINT(5); POINT(6);POINT(7);POINT(8);POINT(9);POINT(10); POINT(11)];

% close all; n = 11; % if (nargin ~= 2) order = 4; % end nplot = 100; % % if (n < order) % display([' !!! Error: Choose n >= order=',num2str(order),' !!!']); % return; % end p = xycoords;

% figure; % hold on; box on; % set(gca,'Fontsize',16);

t = linspace(0,1,nplot);

for i = 1:n % plot(p(:,1),p(:,2),'k-','LineWidth',2); % axis([0 1 0 1]); % hold on; box on; if (i >= order) T = linspace(0,1,i-order+2); y = linspace(0,1,1000); p_spl = DEBOOR(T,p,y,order); % plot(p_spl(:,1),p_spl(:,2),'b-','LineWidth',4); bspliney = p_spl(:,2); bsplinex = p_spl(:,1); controlpoints=p; end % plot(p(:,1),p(:,2),'ro','MarkerSize',10,'MarkerFaceColor','r');

 end

sim 'TESTCOMPLETEwithsimulinkvariance_RRCLEANUP' A = abs((max(response1(:,6)))-(min(response1(:,6)))); B = abs((mean(response1(:,2)))-veldem); C = abs(bspliney(1)-bspliney(end)); x = [A B C]; cost = x(1) + 10000*x(2) + 10000*x(3); assignin('base','cost', cost); end

% global response1

assignin('base', 'response1', response1); end

my DEBOOR is

if true
function val = DEBOOR(T,p,y,order)
% function val = DEBOOR(T,p,y,order)
%
% INPUT:  T     reference points
%         p     control points (nx2-Matrix)
%         y     evaluation points (Spaltenvektor)
%         order spline order
%
% OUTPUT: val   Werte des B-Splines an y (mx2-Matrix)
%
% Date:   2007-11-27
% Author: Jonas Ballani
m = size(p,1);
n = length(y)
X = zeros(order,order);
Y = zeros(order,order);
a = T(1);
b = T(end);
T = [ones(1,order-1)*a,T,ones(1,order-1)*b];

for l = 1:n t0 = y(l); id = find(t0 >= T); k = id(end); if (k > m) return; end X(:,1) = p(k-order+1:k,1); Y(:,1) = p(k-order+1:k,2);

    for i = 2:order
        for j = i:order
            num = t0-T(k-order+j);
            if num == 0
                weight = 0;
            else
                s = T(k+j-i+1)-T(k-order+j);
                weight = num/s;
            end
            X(j,i) = (1-weight)*X(j-1,i-1) + weight*X(j,i-1);
            Y(j,i) = (1-weight)*Y(j-1,i-1) + weight*Y(j,i-1);
        end
    end
    val(l,1) = X(order,order);
    val(l,2) = Y(order,order);
end

end

and what i used to run my ga was

if true
  % global POINT
global tstep
global veldem
global bspliney
global bsplinex
global response1
global controlpoints
global xycoords
global ycoords
global cost
veldem = 52;
POINT = zeros(1,11);
POINT(:)=veldem;
tstep=0.001;
%  gt=mycostfunction(POINT);
x0 = zeros(1,11);
x0(:) = veldem;
[x,fval,exitflag] = fmincon(@mycostfunction,x0);
dbstop if error
end

can someone please help me, I also tried running this with fmincon and although it worked, only initial values were sent back to me?

My simulink just models a four bar mechanism and returns torque which i am trying to minimise the peak to peak of

any help is greatly appreciated thank you in advance