In this paper, a trajectory tracking control strategy based on radial basis function neural network (RBFNN) learning control is proposed for servo systems. A multi-parameter adaptive RBFNN is first constructed with update rate design. Subsequently, weight generalization rules are introduced to achieve high-precision tracking of test trajectories based on learned knowledge without retraining. Furthermore, for trajectory magnification scenarios, a specific generation algorithm is designed, which is combined with the generalization rules and achieves high-precision tracking of magnified trajectories that are not included in the training data. Experiment results demonstrate that the proposed method can achieve high-precision tracking of test trajectories and generalization to similar trajectories.