Abstract: Taking a vehicle swing arm rubber bushing as the research carrier, the axial frequency correlation and amplitude correlation of the bushing are studied through the dynamic loading test. A rubber bushing model based on parallel connection of an elastic unit, a friction unit and a higher-order fractional derivative viscoelastic unit is established. Based on the dynamic neighbor and generalized learning particle swarm optimization algorithm, a higher-order fractional derivative viscoelastic unit parameter identification method is proposed. Then the model parameters are identified combined with the test results. Furthermore, the fractional derivative coefficients are fitted to the functions of the loading amplitudes, which improves the overall prediction accuracy and applicability of the model. It is verified by test results that the established model can accurately describe the axial dynamic characteristics of the suspension rubber bushing. The model is of great significance for improving the accuracy of dynamic simulations of suspension system and the vehicle

Key words: rubber bushing, dynamic characteristics, fractional derivative, particle swarm optimization, parameter identification