'Not enough input arguments' when I am defining a function

Hi,

I am trying to link SimEvent and the optimization module of MATLAB. For that, I first need to define a function that runs the simulation then call it in the optimization function that I will define later. But I am getting the "not enough arguments" when defining the function. And I cannot fix it! here is my code and the warning:

 function obj = SimOpt(vecX)

        NumServers = vecX(1);
        NumTruck = vecX(2);

        set_param('concreting10/Positioning and Unloading','NumberOfServers',num2str(NumServers));
        set_param('concreting10/Washing','NumberOfServers',num2str(NumTruck));

        simOut = sim('concreting10','SaveOutput','on','OutputSaveName','WaitingTimeInQueue');
        z = simOut.get('WaitingTimeInQueue');

        waiting = max(z);

        cost = [100 200]*vecX';
        obj = waiting*1000+cost;

   end

-----------------------------------------------------------

Not enough input arguments.

Error in SimOpt (line 31) NumServers = vecX(1);

-------------------------------------------------------

Would appreciate any help.

Answer 1

[Walter Roberson]

However it is that your routine SimOpt is getting invoked, it is not being passed any inputs. How is your routine getting invoked?

[Zahra Moussavi]

Thanks Walter for your answer.

As I said this function is part of another one. Let me put all the code;

unction finalresults = SimOpt ()

intcon = [1];

A=[];

b=[];

Aeq=[];

beq = [];

lb = [1];

ub= [10];

inalresults= intlinprog(@f,intcon,A,b,Aeq,beq,lb,ub);

    function obj = f(vecX)

        NumServers = vecX(1);
        NumTruck = vecX(2);

        set_param('concreting10/Positioning and Unloading','NumberOfServers',num2str(NumServers));
        set_param('concreting10/Washing','NumberOfServers',num2str(NumTruck));

        simOut = sim('concreting10','SaveOutput','on','OutputSaveName','WaitingTimeInQueue');
        z = simOut.get('WaitingTimeInQueue');
        waiting = max(z);

        cost = [100 200]*vecX';
        obj = waiting*1000+cost;

    end
end

------------------------------------------------------

When I run the whole code I get this warning:

Error using intlinprog (line 122) INTLINPROG requires the following inputs to be of data type double: 'f'.

Error in SimOpt (line 26) finalresults= intlinprog(@f,intcon,A,b,Aeq,beq,lb,ub);

Answer 2

[Alan Weiss]

I think that you have a misunderstanding about what intlinprog does. It does NOT minimize nonlinear functions, such as a function that is the result of a simulation. It minimizes a LINEAR function of a variable x, such as

 f(1)*x(1) + f(2)*x(2) + ... + f(N)*x(N)

where the f vector is a constant numeric vector. You appear to be attempting to minimize a nonlinear function by passing a function handle @f, and that is why you get an error.

It is possible that you are trying to minimize a nonlinear function while restricting x(1) to be integer-valued. If that is the case, then I recommend that you minimize using fmincon or fminunc with the value of your x(1) variable set manually to various integer values. You can easily program a search over a one-dimensional integer range, especially since it looks like you want x(1) to go from 1 to 10.

Alan Weiss

MATLAB mathematical toolbox documentation

[Zahra Moussavi]

Thanks Alan for your answer.

Well I got the idea of the simulation/optimization code from the link below:

http://au.mathworks.com/videos/optimizing-manufacturing-production-processes-68961.html

I tried to go through all the code I see in this video but, when I am applying it, it is not working.

[Zahra Moussavi]

Now that I am looking at it again, I see that in the video Genetic Algorithm was used as the optimization toolbox. Will try to use that. Hopefully I can make it work.

Thanks again