Abstract : Based on the traditional PID controller, a neural network controller suitable for CNC machine tool position servo control is proposed based on the self-learning characteristics of neurons. This algorithm is characterized by the absence of system modeling, simple structure, real-time adaptive adjustment of control parameters, simple computation, and applicability to practical engineering applications. Simulation and real-time control experiments of CNC machine tool servo systems show that the adaptive control algorithm remains effective even when the operating environment of the controlled object is unknown. Keywords : Computer numerical control; Position servo system; Neural network; Adaptive control The maximum speed, tracking accuracy, and positioning accuracy of CNC machine tools depend on the dynamic and static performance of the drive and position control systems. Therefore, researching and developing high-performance drive and position control systems has always been one of the key technologies in CNC machine tool research. Currently, classical control methods, such as proportional or proportional-integral algorithms, are still widely used in CNC machine tool position servo control due to their simplicity and ease of implementation. However, it suffers from drawbacks such as poor adaptability of control parameters and weak anti-interference ability. To meet the manufacturing industry's pursuit of high-efficiency production of high-quality products and the need to process increasingly complex parts, it is necessary to continuously improve and enhance the steady-state accuracy, dynamic response characteristics, adaptability to system parameter changes, and anti-interference ability of position servo systems. Therefore, adopting and developing advanced control technologies is an inevitable trend. However, many currently proposed control algorithms still have significant gaps compared to practical applications, mainly in the following aspects: ① Limited by the computational load of the algorithms, it is difficult to meet the real-time control requirements; ② The control theory is imperfect in parameter design and stability analysis; ③ Modeling errors limit the control quality. This paper utilizes the self-learning function of neural networks to design an online intelligent position controller and applies it to the real-time control of actual CNC machining, achieving good control results. [b][align=center]For details, please click: CNC Machine Tool Neuron Adaptive Position Control Algorithm[/align][/b]