Simulation study on assist-as-needed control of a rehabilitation robotic walker

BACKGROUND: Along with China entering an aging society, the percentage of people that over 60 will reach 34.9% in 2050, resulted in a significant increase in stroke patients. OBJECTIVE: This paper proposes a rehabilitation robotic walker for walking assistance during the daily life, and a control method for the motor relearning during the gait training. The walker consists of an omni-directional mobile platform (OMP) which ensures the walker can move on the ground, a body weight support system (BWS) which is capable of providing the desired unloading force, and a pelvic assist mechanism (PAM) to provide the user with four degrees of freedom and avoid the rigid impact. The study goal is to gain a better understanding of the assist-as-needed control strategy during the gait training. METHODS: For the man-machine interaction control, the assist-as-needed control strategy is adopted to guide the users’ motions and improve the interaction experience. To build the force field in the three-dimensional space, the dynamics of the system is derived to increase the accuracy of force control. RESULTS: The simulation results show that the force field around the motion trajectory was generated in the three-dimensional space. In order to understand the force field, we designed the simulation on sagittal plane and the controller can generate the appropriate force field. The preliminary experiment results were consistent with the simulation results. CONCLUSION: Based on the mathematical simulation and the preliminary test, the results demonstrate that the proposed system can provide the guide force around the target trajectory, the accuracy of force control still remains to be improved.