A wearable elbow exosuit system has been proposed in this work, including the origami-inspired exosuit structure along with a portable air source and electromyography (EMG)-based active rehabilitation control method. The elbow exosuit is designed using an origami-inspired pneumatic actuator to meet the biomechanic requirements for elbow assistance. And a portable pneumatic source attached to the waist is also proposed to drive the elbow exosuit. On the basis of exosuit structure design, the active control with cascaded frame is then developed. For the active perspective, the EMG-based motion prediction is accomplished for the input of controller. To achieve real-time and accurate prediction, a simple feedforward neural network is utilized for a motion prediction model based on its fast training. To further reduce the size of the network, the features are extracted from the EMG and angle for the inputs, replacing the end-to-end method. Based on intention prediction, the cascaded controller subsequently completes position control, torque control and pressure servo control. Finally, through preliminary experiments on healthy participants, the elbow can be accurately predicted for the EMG-based method, and the assistance efficiency is verified through task scores and reduction in muscle activation. In summary, the proposed wearable exosuit can provide a reference for the design of wearable devices.
Chen et al. (Tue,) studied this question.