This paper aims to reduce the heating of the electromagnetic actuators of a magnetic bearing. The electrical current of the coils was above normal, so the need for a new coil design to reduce heating due to high currents. In this scenario, the proposed methodology allows minimizing copper losses using Particle Swarm Optimization, so that the best result of the design parameters will be used in the construction of the new coil for the actuator. For the development of this work, it was decided to use a computational tool for public use, FEMM (Finite Element Method Magnetics) to simulate the electromagnetic device. In the simulations, the densities of magnetic fluxes in the core and in the air gap are shown, as well as the energies, electromagnetic forces and losses in the copper of the electromagnetic actuator winding. Finally, an optimal model of the actuator is obtained through the use of optimization techniques. Therefore, the results obtained demonstrate that the proposed methodology is configured as an interesting strategy for the purpose of this work.