Research on Control Method of Arm Rehabilitation Robot Based on BP Neural Network

Rehabilitation robot is a kind of human-computer interaction robot, and its motion mode must conform to the motion characteristics of patients' limbs in recovery period and meet the training requirements of patients in different recovery periods. The arm rehabilitation robot is a bionic manipulator which takes the patients with arm hemiplegia as the object and carries out a series of rehabilitation training based on rehabilitation medicine. Patients with arm hemiplegia only need to wear the arm rehabilitation robot on the affected limb, and then choose passive exercise, active auxiliary exercise and active damping training methods according to the recovery of their own arms, and complete daily rehabilitation training independently. For the rehabilitation treatment of upper limb dysfunction, it is very important to adhere to long-term and effective scientific treatment. Compared with the traditional methods, BP neural network has faster convergence, higher learning accuracy and better network generalization ability. According to the structure of the arm rehabilitation robot, this paper discusses the control method of arm rehabilitation robot based on BP neural network from the perspective of intelligent control.