Genetic Algorithms (GAs) are a technique that has given good results to those problems that require a search through a complex space of possible solutions. A key point of GAs is the necessity of maintaining the diversity in the population. Without this diversity, the population converges and the search prematurely stops, not being able to reach the optimal solution. This is a very common situation in GAs. This paper proposes a modification in traditional GAs to overcome this problem, avoiding the loose of diversity in the population. This modification allows an exhaustive search that will provide more than one valid solution in the same execution of the algorithm