In view of the synchronization error caused by the characteristics of motor and the change of load, the dual motor drive system of the automatic feeding mechanism of the press machine is studied in this paper. The dynamic characteristics of the servo motor were analyzed, and the factors that affect the synchronous control were found out to provide the basis for designing the synchronous control scheme. A single neuron PID control strategy was proposed to adjust the PID parameters in real time by changing the weight coefficients of the neural network, so as to solve the problem that the traditional PID cannot be adjusted online. At the same time, on the basis of the cross coupling control method, the simulation analysis and experiment were carried out with the feed speed of the screw, the position error and the control current of the servo motor as the control index. The research results show that compared with the traditional control method, the proposed method can greatly improve the synchronous control precision of two motors, effectively reduce the synchronization error between the screw, and the antiinterference ability is strong, and it could meet the design requirements.