Date of Award
8-2026
Document Type
Thesis
Degree Name
Master of Science (MS)
Department
Mechanical Engineering
Committee Chair/Advisor
Dr. Ge Lv
Committee Member
Dr. Joshua Bostwick
Committee Member
Dr. Divya Srinivasan
Abstract
Gait phase is a human-inspired physical variable that describes the instantaneous position of a person within a gait cycle. Adaptive oscillators have been widely used for gait phase estimation due to their ability to synchronize with periodic and pseudo-periodic biomechanical signals and generate monotonically increasing cyclic phase values. Typically, the performance of adaptive oscillators is strongly influenced by coupling strength, a parameter that bridges a trade-off between convergence speed and steady-state variability of the estimated stride frequency. In gait assistive applications, stride frequency estimation must be sufficiently fast to enable phase-based assistance while achieving a stable stride frequency estimate. The objective of this study is to derive an adaptative dynamicsfor coupling strength, that would achieve minimum stride frequency convergence time and accurate gait phase estimation, across constant speed walking, speed transitions and task transitions. Additionally, due to an inherent finite convergence time required for adaptive oscillators to synchronize with changing task-dependent stride frequency during locomotion task change, phase estimation becomes unreliable during transitionary periods. This unreliability on gait phase estimation is monitored using a task transition detector, which identifies the detection of task transition onset and transient synchronization behavior as the adaptive oscillator converges toward task-dependent stride frequency.
Recommended Citation
Radha Krishnan, Nikhila, "Extended Adaptive Oscillator Framework for Gait Phase Estimation with Adaptive Coupling and Transition Detection across Locomotion Tasks" (2026). All Theses. 4819.
https://open.clemson.edu/all_theses/4819