The dynamic stiffness of traction motor bearings directly affects the dynamic performance and operational stability of rail transit rotor systems.To address the limitations of existing testing methods in balancing multi-condition measurement and nonlinear feature extraction,this paper proposes a data-and model-driven method based on measured data and theoretical models for high-preci-sion testing of rolling bearings'dynamic stiffness.Firstly,based on the analytical theory of the rolling bearing stiffness matrix,a five-de-gree-of-freedom quasi-static model was established to systematically consider nonlinear factors such as centrifugal force,gyroscopic mo-ment,and contact deformation.Secondly,a dynamic stiffness test rig for bearings was independently developed to achieve high-precision synchronous measurement of bearing displacement and applied force under continuously variable load conditions.Thirdly,a parameter optimization objective function integrating measured data and theoretical models was constructed,and an intelligent optimization algo-rithm was employed to identify key parameters defining the bearing structural characteristics.Finally,the accurate solution of the bear-ings'dynamic stiffness was obtained through the analytical model.Experimental results indicate that the bearings'dynamic stiffness ex-hibits a significant"stiffness softening"phenomenon with increasing speed;the axial dynamic stiffness increases markedly with increas-ing axial load,while the radial dynamic stiffness shows a relatively weak trend of variation with increasing radial load.The proposed method effectively overcomes the traditional linear fitting's limitation of ignoring nonlinear characteristics,providing an important theo-retical basis and experimental support for the dynamic design and operational stability assessment of traction motor rotor systems.