Linear constraints in FMINCON Optimization

I'm using fmincon for solving an optimization problem. Mathematical model consists of 12 equations, control vector u contains 4 elements.

State vector:

x =[0 3 0 0 5 0 0 0 12 0 0 0];        

I want to put linear constraints on the on the 2nd, 5th and 9th elements, so that x(2)=<5, x(5)=<10 x(9)>=10. I wrote it in the following way:

function [A, b, Aeq, beq, lb, ub] = linearconstraints(t, x, u)
    A   = [0,1,0,0,0,0,0,0,0,0,0,0;
          0,0,0,0,1,0,0,0,0,0,0,0;
          0,0,0,0,0,0,0,0,0,1,0,0]; 
    b   = [5;10;-10]; 
    Aeq = [];
    beq = [];
    lb  = [];
    ub  = [];
end

During the simulation Matlab shows the error:

Error using fmincon (line 298). A must have 4 column(s).

I assume fmincon requires the formulation of A, corresponding to the number of control inputs u (4).

Can you recommend how should the A and b look like so fmincon will work correctly?

Thank you in advance for your answer!

[Torsten]

You have to put all unknown variables in one vector. In your case, this is most probably the state vector + the control vector, thus a vector of length 16. And this will also be the size of your initial vector when calling "fmincon".