Re: [Help-glpk] Can GLPK for JAVA solve multi-objective optimization problem?

[Xypron]

Hello Yi,

In your optimization problem quadratic terms my(k) * my(k) appear. GLPK 
can only solve linear problems.

Hence if you want to use GLPK you will have to change the formulation a 
bit. Below I give a description in GMPL which you could use with the 
standalone solver glpsol:

param H; # replaces H_p - 1
set K := {0..H};
set I := {0..10};
param c{K};
param t{I}; # t replaces the term (T_in[K] - T_ref)^ 2
var  x[K,I], binary;

minimize obj :
  sum{k in K, i in I} c[k] * i * x[k,i] + sum{k in K, i in I} x[k,i] * 
t[i];

s.t. c1{k in K}
  sum{i in I} x[k,i] = 1;

solve;

printf {k in K} "my[%d] = %d\n", k, sum{i in I} i * x[k,i];

end;


GLPK for Java comes with a file examples/Mip.java showing how to create 
a MIP problem in Java.

For details of the methods and constants used consult doc/glpk.pdf of 
the GLPK source distribution.

Best regards

Xypron

Zong, Yi wrote:
> Hi!
>
>       I have an objective function described as following:
>
>      The optimized variables are , which are integer between [0, 10], 
> is a lineal function of  an
>
> , and
>
>    Can I use GLPK for JAVA to get the optimized results? If it is 
> possible, could you give me an example?