Articles techniques et présentations

In this paper a poly-optimization of the design of the electromechanical actuator is presented. The shape of the actuator is defined by the decision variables. The number of decision variables under consideration is up to ten but in the next step while the multi-coils system will be analyzed the number of decision variables will increase up to hundred, so the genetic algorithm has been used. The genetic algorithm program has been implemented in the Matlab program which operates together with the COMSOL Multiphysics® package. The aim of the optimization process is to find the maximal electromagnetic force and the minimal device mass. The final result is a set of the Pareto optimal solutions, which makes possible to draw out some more general conclusions on a design of the actuator.