functioncplexexpln

Does Cplex support using of ln or exponential function for a decision variable?

I am trying to write constraints using ln and exp function, yet I received an error that Cplex can't extract the expression.

forall (t in time)
   Gw_C["Mxr"] == 20523 + 17954 * ln(maxl(pbefore[t]));

  Ed_c ["RC"]== 0.0422* exp(0.1046* (maxl(pbefore[t])));
   
   Gw_C["RC"] == 3590* pow((maxl(pbefore[t]), 0.6776);

Is there any other possible way to code these constraints on cplex? Thanks

Solution

  • You may use exp and log if you rely on Constraint Programming within CPLEX:

    using CP;

int scale=1000;

dvar int scalex in 1..10000;
dexpr float x=scalex/scale;

maximize x;

subject to
{
  exp(x)<=100;
}

execute
{
  writeln("x=",x);
}

    works fine and gives:

    x=4.605

    But with Math Programming within CPLEX you cannot use exp like that.

    What you can do instead if go through linearization.