Control strategy and experimental study of brushless DC motor based on fuzzy BP neural network

Traditional PID control suffers from issues such as low control accuracy, significant overshooting, and poor disturbance rejection capabilities. This study explores a hybrid control strategy for a brushless DC motor using a combination of fuzzy backpropagation (BP) neural network and proportional-integral-derivative (PID) control. The aim is to enhance the precision and adaptability of the motor control system. The fuzzy BP neural network is employed to address the inherent nonlinearity and uncertainty in the motor system, thereby improving control performance. PID controllers contribute to increased system stability and rapid response capabilities. A simulation control program is constructed in the Matlab/Simulink environment to place the brushless DC motor under various operating conditions. Fuzzy PID control and fuzzy BP neural network PID control methods are applied separately to control the brushless DC motor, evaluating the effectiveness of each control approach. The research results validate that the brushless DC motor controlled by the proposed fuzzy BP neural network algorithm achieves a final speed closer to the target speed, with lower overshooting and speed error, reduced torque fluctuations, and better adaptation to environments with significant disturbances. The control strategy exhibits good disturbance rejection and robustness, effectively improving the overall dynamic performance of the motor control system.