Skip to content
Open access

Task-Space Control for Powered Knee–Ankle Prostheses

2026 · IEEE transactions on neural systems and rehabilitation engineering · Vol 34, pp. 4347-4358 · 0 citations · 77 references

Abstract

As the capabilities of powered lower-limb prostheses continue to improve, there is an opportunity to shift the focus of prosthetic control beyond purely joint-level considerations and incorporate whole-body task-based information to better coordinate user and device interaction. This paper presents a control framework, termed task-space control (TSC), that determines the desired torque response of a knee-ankle prosthesis during level-ground walking and stair ascent based on task-level objectives, specifically ground reaction forces (GRFs) and center-of-mass (CoM) kinematics. TSC is evaluated with two individuals with knee disarticulation. While results were subject-specific, relative to a phase-based impedance controller, TSC with CoM feedback reduced stair-ascent gait cycle duration for one individual by 17.6%, and reduced stair-ascent CoM tracking error for the other by 27.6% vertically and 32.4% horizontally. More broadly, while exclusively considering task-level objectives to compute actuator torques, TSC still produced key joint-level kinematic outcomes for multi-joint and multi-modal control, like plantarflexion push-off in late stance for level-ground walking and powered knee extension for step-over-step stair ascent. However, subjects did not prefer certain normative characteristics, like early stance knee flexion during level-ground walking. These preferences raise questions about whether fully replicating normative joint characteristics should be the primary target for prosthetic control, or if system-level approaches like TSC may be better suited for adapting prosthesis behavior in conjunction with the user.

Read PDF

We use cookies to run the site and, with your consent, for analytics and to show ads. See our Cookie Policy.