The study and understanding of algorithms that solve combinatorial problems based on swarm intelligence continuous metaheuristics, is an area of interest at the level of basic and applied science. This is due to the fact that many of the problems addressed at industrial level are of a combinatorial type and a subset no less than these are of the NP-hard type. In this article, a mechanism of binarization of continuous metaheuristics that uses the concept of the percentile is proposed. This percentile concept is applied to the An Lion optimization algorithm, solving the set covering problem (SCP). Experiments were designed to demonstrate the importance of the percentile concept in the binarization process. Subsequently, the efficiency of the algorithm is verified through reference instances. The results indicate that the binary percentile gravitational search algorithm (BPGSA) obtains adequate results when evaluated with a combinatorial problem such as the SCP.