High-end CNC (Computer Numerical Control) machine tools may encounter control problems due to nonlinear friction during low-speed motion. In order to improve the motion accuracy and stability of the machine tool, this study proposes a parallel hybrid control model based on BP neural network (Back Propagation Neural Network) and PID (Proportional Integral Derivative) algorithm. By utilizing the adaptive learning ability of BP neural network and the stability of PID algorithm, the optimization control of machine tool motion is achieved through parallel hybrid control strategy. Compared with traditional control models, in terms of control accuracy, the hybrid control model maintains 100% accuracy at 950RPM, while the accuracy of the traditional model decreases to 92.8%. In terms of overshoot, the hybrid control model has an overshoot of 13.6% in the low-speed range (500–600 RPM), while the traditional PID control has an overshoot of 18.5% under the same conditions. These data indicate that the hybrid control model plays a prominent role in the operation of high-end CNC machine tools due to its combination of PID control and BP neural network advantages.