Abstract: Aiming at the parameter optimization design problem of spherical bearingless flywheel machine, a multi-objective stratified optimization design scheme based on parameter sensitivity analysis is proposed. On the basis of introducing the machine operation mechanism and electromagnetic analysis, taking torque and suspension force as the optimization goals, the machine parameters are divided into main sensitivity parameters and sub-sensitivity parameters by constructing sensitivity equations, and support vector machines are used for nonparametric modeling for primary sensitivity parameters and subsensitivity parameters, and optimization is carried out by chaotic particle swarm optimization with adaptive change of inertial weights. Finally, the effectiveness of the proposed algorithm is verified by finite element simulation, and the results show that the machine torque is increased by 6% and the suspension force is increased by 27.99% after optimization.