How could I make a Boolean chromosome with Genetic algorithm

How could I make the fitness function for the GA individuals vector or chromosome.

How could I make the result of the GA equal to ''1 or 0''.

How could I add or returned the fitness value to the GA in order to get goodness.

Answer 1

[Walter Roberson]

The fitness function should seldom return 0 or 1; it should return a value which increases as the proposed X gets further away from optimum.

To have a boolean variable, set its minimum to 0, its maximum to 1, and include it in the IntCon (integer constraints) list.

[Ibrahim Ibrahim]

Thanx very much Dr. Walter for your answer. It was very helpful.

Please could you help me with:

1. How could I make the fitness function for the GA individuals vector or chromosome.
2. How could I add or returned the fitness value to the GA in order to get goodness.

[Walter Roberson]

Example:

   function fitness = myfitness(X)
     Xnum = X(1:2);   %some of the variables are continuous
     Xlog = logical(X(3:5));   %some of the variables are logical

     %now build some potential contributions that might be turned on or off
     parts = [Xnum(1).^2 - Xnum(2).^3, ...
              cos(Xnum(1)) * abs(gamma(Xnum(2))), ...
              abs(sqrt(Xnum(2)) ];
     activeparts = parts(Xlog);   %extract the parts corresponding to variables turned on

     fitness = sqrt( sum(activeparts.^2) );  %combine the contributions somehow

This make some calculations, decides which parts it wants to use according to the boolean inputs, extracts those contributions, and then combines the contributions. The sqrt() of the sum() of squares is not the only valid way to combine the contributions: I just picked some formula for the example.