Data-Driven Variable Impedance Control of a Powered Knee-Ankle Prosthesis for Adaptive Speed and Incline Walking

• DOI: 10.36227/techrxiv.19165895.v2
• Publication Date: 2022-08-11T05:13:50Z
• Authors: Thomas Best; Cara Welker; Elliott Rouse; Robert Gregg
• Keywords: Cadence; State variable; Benchmark (surveying); Impedance Control

<p>Most impedance-based walking controllers for powered knee-ankle prostheses use a finite state machine with dozens of user-specific parameters that require manual tuning by technical experts. These are only appropriate near the task (e.g. speed and incline) at which they were tuned, necessitating many different parameter sets variable-task walking. In contrast, this paper presents data-driven, phase-based controller uses continuously-variable impedance control during stance kinematic swing to enable biomimetic locomotion. After generating data-driven model variable joint convex optimization, we implement novel task-invariant phase real-time estimates incline autonomous adaptation. Experiments above-knee amputee participants (N=2) show our 1) features highly-linear accurate estimates, 2) produces kinetic trends as varies, leading low errors relative able-bodied references, 3) work cadence varies. We presented meets often exceeds performance benchmark two participants, without requiring tuning. </p>