Approach to Gait Coordination: Adaptive Fuzzy Finite-Time Control of a Stochastic Prosthesis-Human Symbiosis with Intentional Delay

• Published in: IEEE transactions on fuzzy systems Vol. 31; no. 11; pp. 1 - 15
• Main Authors: Ma, Xin, Zhang, Xiaoxu, Fang, Hongbin, Xu, Jian
• Format: Journal Article
• Language: English
• Published: New York IEEE 01.11.2023; The Institute of Electrical and Electronics Engineers, Inc. (IEEE)
• Subjects: Adaptive control; adaptive fuzzy finite-time control; Coordination; Delay; Delays; Fuzzy control; Fuzzy logic; fuzzy logic system (FLS); Fuzzy systems; Gait; Human engineering; Human-robot interaction; intentional delay; Legged locomotion; Liapunov functions; Prostheses; Prosthetics; Robots; Statistical analysis; Stochastic processes; Stochastic prosthesis-human symbiosis; Symbiosis; Walking
• ISSN: 1063-6706; 1941-0034
• DOI: 10.1109/TFUZZ.2023.3270707

The generation of intentional delay in response to the stride frequency is seldom considered in prosthesis-human symbiosis. Unfortunately, such intentionally delayed human-robot interaction poses a new challenge to their gait coordination in stochastic environments. Utilizing fuzzy logic systems (FLSs), we investigate an adaptive fuzzy finite-time control of a stochastic prosthesis-human symbiosis with intentional delay to address this issue. Noting that the intentional delay is related to walking velocity, this work conducts experiments on ten healthy subjects to identify the intentional delays at different velocities using the FLS. Introducing the FLS-identified delay and contralateral healthy limb gaits, we propose a prosthetic gait planner to simultaneously determine the reference stride frequency and stride length, thus properly regulating the desired velocity. Considering the adverse effects of the required intentional delay and state constraints in the stochastic framework, we propose a new statistical Lyapunov-Krasovskii functional, together with a Tan-type barrier Lyapunov function. Correspondingly, an adaptive fuzzy controller is developed via a backstepping design, thus solving the semi-global finite-time stable in probability with intentional delay, unknown nonlinearities, and state constraints. Application studies validate the efficacy of the proposed approach. The results show that our approach can predict walking behavior while performing gait coordination.