Research on Series Elastic Damping Actuator Control System Based on BP Fuzzy Neural Network PID Control

The series elastic damping actuator (SEDA) is a crucial mechanical configuration in lower limb exoskeleton rehabilitation robots. It is based on the utilization of elastic elements to enable the system to adapt to external changes and impacts to a certain extent. However, maintaining the stability of the driver is essential during its usage. Addressing the issue of vibrations in SEDA control, this study proposes a hybrid control method that combines fuzzy control, back propagation (BP) neural networks, and proportional-integral-derivative (PID) control. This hybrid approach forms a BP fuzzy neural network PID control method. A network model is established, and MATLAB is employed for simulation analysis. A comparison is made against traditional PID control and fuzzy PID control. The results indicate that the BP fuzzy neural network PID control method achieves the highest precision, ensuring a stable system with no overshooting. Moreover, it exhibits the fastest dynamic response speed when subjected to disturbances.