A Human Joint Torque Estimation Method for Elbow Exoskeleton Control

Abstract

Exoskeleton for motion assistance has obtained more and more attention due to its advantages in rehabilitation and assistance for daily life. This research designed an estimation method of human joint torque by the kinetic human–machine interaction between the operator’s elbow joint torque and the output of exoskeleton. The human elbow joint torque estimation was obtained by back propagation (BP) neural network with physiological and physical input elements including shoulder posture, elbow joint-related muscles activation, elbow joint position, and angular velocity. An elbow-powered exoskeleton was developed to verify the validity of the human elbow joint torque estimation. The average correlation coefficients of estimated and measured three shoulder joint angles are 97.9%, 96.2%, and 98.1%, which show that estimated joint angles are consistent with the measured joint angle. The average root-mean-square error between estimated elbow joint torque and measured values is about 0.143[Formula: see text]N[Formula: see text]m. The experiment results proved that the proposed strategy had good performance in human joint torque estimation.

Artificial Intelligence-Based Wearable Robotic Exoskeletons for Upper Limb Rehabilitation: A Review

“…These characteristics have allowed caregivers to implement different rehabilitation protocols for elderly populations or patients with motor impairments. Other devices are less compact but provide high levels of support [ 42 , 48 , 53 , 54 , 55 ]. For the system by Bakri [ 42 ], simulations and resistance tests were carried out to confirm the proper functioning of the material with multiple weights on the mechanical structure.…”

“…Complex systems have implemented an active operation mode [ 27 , 32 , 41 , 48 , 50 ], which was focused on providing full assistance to the affected limb. The support to the upper limbs is triggered by the measurement of the activity in the electroencephalography (EEG) or electromyography (EMG) systems, and in some cases by eye-tracking data.…”

“…In general, these developments are able to assist the movement of the joint for both flexion and extension movements. In [ 48 ] it was shown that the two fixed forearm blocks slid on a linear rail to provide the forearm DoF, allowing the exoskeleton’s joints to synchronise with the patient’s joints.…”

“…Using a similar approach, the articles [ 22 , 43 , 48 ] presented intelligent strategies with the addition of electromyographic (EMG) and force-myographic (FMG) signal processing, among others. The use of a simple ANN with backpropagation algorithm learning (BP) was proposed in these cases, which led to the development of discrete and continuous motor control models for exoskeleton motion.…”

“…Once the BP-ANN was trained, predictions were made with unlabelled test data. Despite successful results, the implementation of fuzzy systems to reduce the training time of BP-ANN was highlighted as a further improvement [ 48 ].…”