Path Planning for Upper Limb Rehabilitation Based on Human Motion Feature Output

In the field of rehabilitation robot path planning, the trajectory of rehabilitation training is the premise and foundation of rehabilitation robot control. However, the existing movement trajectory generation methods still have many shortcomings, such as the accuracy of data collection and the effectiveness of optimization algorithms, and rarely consider the laws of rehabilitation medicine. In this paper, the human movement data during the upper limb rehabilitation training is collected through a three-dimensional movement capture system. According to the processed data, find the output of human upper limb movement feature and fit it as an output function to obtain the initial training path of the upper limb rehabilitation training. Adjust the weight of the index to obtain the optimal rehabilitation training path focusing on training effect or comfort by setting the rehabilitation training index. The golden section algorithm is used to calculate the upper limb output data constrained by the rehabilitation training index, and the final human body norm output function is obtained through Fourier fitting. The simulation and results show that the canonical output function improves the efficiency of rehabilitation training and ensures the effectiveness and comfort of upper limb rehabilitation.