In order to accurately describe the nonlinear characteristics between process parameters and the finished surface properties after ultrasonic strengthening, the optimal parameters of the spindle speed, the tool head amplitude and the number of extrusion were selected in the ultrasonic strengthening process.According to the principle of grade, based on 28 sets of data were obtained of feed velocity and pressure. Surface roughness and hardness data was obtained by experimental measurements. The resulting datas were used into the nonlinear hysteretic model of support vector machines and the roughness and hardness nonlinear characteristic model of ultrasonic strengthening was established.The real data was using to verify the model finally. The results show that the predicted data coincides with the original data.It is proved that the model is feasible for the prediction of the surface roughness and hardness of the parts after ultrasonic machining.