if there are constraints, then the fitness function value doesnt equal to the objective function value --> you will have to account for the constraints or the -ve sign in the question (but eventually when you get the final solution, the -ve of the fitness function will be equal to the obj function, under the same optimal solutions)... however, when you get solutions that do not meet the constraints, then the penalty function will kick in, and render the fitness function value to be a lot smaller than the obj function value (cos fitness = -obj fn - penalty - penalty)