Design optimization of permanent magnet motors by evolution strategies and finite element analysis

The power and force density of electric motors becomes higher and higher and it is very important to design most optimal machines. Conventional methods don't reach this aim. A new approach is presented combining evolutionary strategies (ES) and finite element analysis (FEA) to obtain reliable results. The task of our research project is to develop in-wheel motors. An electric motor is directly connected to a wheel without a gearbox making possible very compact, quiet and robust drives. This requires high torques at low speeds. Permanent magnet motors with a high number of poles and an outer rotor carrying magnets are most suitable for this application. The relationship between the geometrical and electrical parameters are nonlinear and highly complex. Therefore a universal applicable procedure, the evolution strategy, is used. The optimization has to find the minimum or maximum of the objective function presenting the best solution. This function may include the motor torque, the efficiency, the costs or other or a combination of them. For estimating the latter a finite element analysis is used to receive accurate results and to consider nonlinear material properties. This paper shows in particular the adaptations of evolution strategy required to get best results in such optimizations.