The work presented in this thesis focus on electrical machines based on flux-switching principle. We were studying topologies with permanent magnets, DC field coils or hybrid-excited (combining both PMs and DC coils) structures. In the first part of this manuscript, we are meeting industrial needs being set by our partner Leroy Somer. The aim is to design in an optimal way a low cost “DC Generator”. Numerical simulations of unconventional hybrid-excited structures are firstly proposed. Then, to carry out a global optimization of those topologies, the numerical model is coupled with a genetic algorithm. The second part of this work is dedicated to an analytical model to predict the airgap field in conventional and unconventional switched-flux machines with DC coils. Two approaches are proposed, either using the Magnetomotive force- Permeance theory or directly solving the field governing equations in the doubly-slotted airgap of those structures. After, an analytical model to determine the magnetic field in ferromagnetic parts is proposed. This model can account for bidirectional field in stator and rotor yokes. This model can be used a popsteriori to assess iron losses in the core. Finally, influences of the number of rotor teeth, or iron losses, on optimal electromagnetic performances are investigated using a stochastic algorithm. In the third and last part of this thesis, we detailed an original approach named topological exploration. We first derivate an analytical model capable to predict the airgap field in PM excited structure. Then, the topological exploration approach was applied to static PM excited machines.
