From a computational point of view, the identification of Bouc-Wen (BW) hysteresis model is a hard task due to the large number of parameters to be found. This is one of the reasons for which it is rarely used for modeling magnetic hysteresis, where other hysteresis models are widely used (e.g., Preisach and Jiles-Atherton). However, the opportunities that the differential expression of BW model could offer for its use in more complex computation task (e.g., nonlinear inductors inserted into a circuit and so on) justify a deeper investigation on its adoption in ferromagnetism. In this paper, using a new hybrid heuristic called metric-topological-evolutionary optimization (MeTEO), the BW identification is presented. MeTEO is a powerful algorithm based on a synergic and strategic use of three evolutionary heuristics: 1) the flock-of-starlings optimization, which shows not only high exploration capability, but also a lack of convergence; 2) the particle swarm optimization, which has a good convergence capability; and 3) the bacterial chemotaxis algorithm, which has no collective behavior or exploration skill, but has high convergence capability. MeTEO is designed to use parallel architectures and exploits the fitness modification technique. Numerical validations are presented in comparison with the performances obtained using other approaches available in the literature.