Proportional–Derivative Iterative Control Strategy Based on Personalized Gait Trajectory

Abstract The gait trajectory in lower limb rehabilitation devices is a key factor influencing rehabilitation outcomes. Personalized gait trajectory therapy can enhance both training effectiveness and rehabilitation comfort. This study proposes a physiological structural model-based personalized gait trajectory generation method for a plantar-driven rehabilitation robot. To control precisely personalized gait trajectories in training, a proportional–derivative (PD)-type iterative learning control strategy is applied for trajectory tracking. First, a combined kinematic model of the human body and the gait rehabilitation device is established, and this model can generate the personalized gait trajectories. The iterative learning control method is used to ensure accurate trajectory tracking. Second, the feasibility of personalized trajectory generation is verified through simulation to evaluate the effectiveness of the method, the result shows that less data is used compared to previous personalized trajectory generation models. The PD learning rate iterative control well solves the problem of the initial state error due to the different physical characteristics of patients, while the traditional control shows that there is always some error fluctuation. In addition, the effectiveness of the control strategy was verified from aspects such as delayed response and anti-interference. Finally, a prototype was designed and four subjects participated in the experiment, each with four sets of experiments. The experimental results show that the PD iterative control strategy enables personalized gait rehabilitation training for different patients.