A method for distributing reference points uniformly along the Pareto front of DTLZ test functions in many-objective evolutionary optimization