Impedance soft stops for powered knee orthoses: A pilot study of gain-scheduled PD flexion limitation in healthy adults

Abstract

After anterior cruciate ligament reconstruction, adjustable braces commonly restrict knee range of motion using rigid flexion stops, which can produce abrupt impacts and discomfort. Passive compliant mechanisms remain hardware-dependent. We developed a software-defined impedance soft stop for powered knee orthoses, implemented as a gain-scheduled proportional-derivative controller that increases stiffness and damping near a prescribed flexion boundary with angle- and velocity-gated engagement. We conducted a counterbalanced six-condition crossover pilot study (physiological gait, orthosis only, orthosis plus motor with the soft stop disabled, rigid stop, firm soft stop, and mild soft stop) with

healthy adults walking at 3 km/h. Outcomes included knee kinematics from the actuator encoder and optical motion capture, actuator torque and compensatory kinematics, and eight-channel surface electromyography (sEMG). Relative to the motor-only condition, soft stops reduced swing-phase knee flexion (

q 95 $ {q}_{95} $ q 95

) from

61.1 ∘ $ {61.1}^{\circ } $ 61.1 ring

[58.4, 63.8] to

37.1 ∘ $ {37.1}^{\circ } $ 37.1 ring

[36.4, 37.8] (firm) and

44.6 ∘ $ {44.6}^{\circ } $ 44.6 ring

[43.4, 45.8] (mild), and reduced flexion velocity from

3.63 $ 3.63 $ 3.63

to

1.02 $ 1.02 $ 1.02

–

1.47 rad / s $ 1.47\;\mathrm{rad}/\mathrm{s} $ 1.47 rad slash s

. Mean resisting torques remained modest (

2.64 N m $ 2.64\;\mathrm{N}\;\mathrm{m} $ 2.64 Nm

firm,

1.65 N m $ 1.65\;\mathrm{N}\;\mathrm{m} $ 1.65 Nm

mild), and compensatory kinematic changes were generally small. During terminal swing, the analysis identified higher right-biceps-femoris average rectified value with the rigid stop than with both orthosis baselines and both soft-stop conditions. Angle- and velocity-gated impedance soft stops provided effective software-defined flexion limitation in a powered knee orthosis while keeping torque demands low and avoiding increased biceps femoris activation and perceived burden associated with rigid stops. Clinical performance after ACL reconstruction remains to be evaluated.