Torque optimization of a novel reaction sphere actuator based on support vector machines

This paper presents a torque optimization method of a novel reaction sphere actuator for spacecraft attitude control. The proposed reaction sphere is based on a spherical induction motor, which is composed of a stator, 6 electromagnets for magnetic levitation and a rotor made of a multi-layer spherical metal shell. Torque model is built through SVMs and FEM simulations. The optimized structure parameters are calculated using a modified genetic algorithm based on the developed SVM model. The simulation results validate the effectiveness of the proposed optimization method, showing that the proposed reaction sphere with optimized structure could increase the torque density by 57% as compared to the original prototype.