In order to improve the torque and suspension performance and reduce the torque ripple of Spherical Axial Split-phase Permanent Maglev Flywheel Machine (S-AP-MF), a multi-objective optimization design method based on Kriging model was proposed. A joint simulation model of S-AP-MF is constructed using Maxwell and Isight, and the sample data set is obtained. Then, Latin hypercube method is used to select samples evenly and hierarchically to conduct sensitivity analysis on S-AP-MF structural parameters and obtain the variables to be optimized. Based on this, to improve the model accuracy and optimization efficiency, the Kriging model and Neighborhood Cultivation Genetic Algorithm (NCGA) are used to obtain the optimal parameter set of S-AP-MF. Finally, the torque performance and levitation force characteristics before and after optimization are analyzed and compared by Finite Element Method. The results show that the torque and average levitation force of the optimized S-AP-MF are increased by 9.54% and 5.51% respectively, and the torque ripple is reduced by 27.95%, which verifies the effectiveness of the proposed method.