The relationship between physical human-exoskeleton interaction and dynamic factors: using a learning approach for control applications

Abstract

During a human-exoskeleton collaboration, the interaction torque on exoskeleton resulting from the human cannot be clearly determined and conducted by normal physical models. This is because the torque depends not only on direction and orientation of both human-operator and exoskeleton but also on the physical properties of each operator. In this paper, we present our investigations on the relationship between the interaction torques with the dynamic factors of the human-exoskeleton systems using state-of-the-art learning techniques (nonparametric regression techniques) and provide control applications based on the findings. Experimental data was collected from various human-operators when they were attached to the designed exoskeleton to perform unconstraint motions with and without control. The results showed that regardless of how the experiments were done and which learning method was chosen, the resulting interaction could be best represented by time varying non-linear mappings of the operator's angular position, and the exoskeleton's angular position, velocity, and acceleration during locomotion. This finding has been applied to advanced controls of the lower exoskeletal robots in order to improve their performance while interacting with human.

Survey of On-line Control Strategies of Human-Powered Augmentation Exoskeleton Systems

“…(a) BLEEX [3] (b) HULC [16] (c) HAL [17] (d) HUALEX [26] Figure 1: The most famous lower limbs human power augmentation exoskeletons.…”

“…The learning approach of the relationship between physical human-exoskeleton interaction and dynamic factors [26,27] allows efficient application of Admittance Control (AC). Ordinary AC [28,29] can't be applied for maneuverable human-powered augmentation exoskeleton systems ( Figure 1).…”

Bio-Based Supervisory Control of a Lower Exoskeleton for Stance Phase

“…This is fundamentally because the physical interaction forces exerted onto the exoskeleton from the wearer almost change from person to person and also within one person over time. This makes the characteristics of the interaction environment (from human) change [11]. Moreover, motion speed is different at various times, even the motion changes from walking to running.…”

Hierarchical Interactive Learning for a HUman-Powered Augmentation Lower EXoskeleton

“…For high-level motion learning, Dynamic Movement Primitive (DMP) [14], [15] is utilized to model the motion trajectories. In order to update the DMP incrementally, we utilize Locally Weighted Regression (LWR) [10], [11] with demonstrated motion trajectories via physical human-exoskeleton interaction. For low-level controller learning, reinforcement learning [12], [13] is employed to let the exoskeleton behave according to the reshaped motion trajectories.…”