Bipedal locomotion is a routine and yet sophisticated task. Studies looking at biological systems suggest that bilateral coordination is the key to articulated rhythmic motions such as walking. Given its importance in formulating human walking, the potential of promoting interlimb coordination to facilitate gait-training for post-stroke patients who suffer impaired motor function due to neurological lesions, has NOT been fully explored. Current strategy primarily focused developing a more normal swing pattern on the affected side, but the results had mixed success so far. The gait-training community has yet to close the loop between prescribing the appropriate interventions and evaluating corresponding changes on either biomechanical or neurological domain.

Inspired by a series of reduced-order locomotion models, this thesis aims at proposing a model framework, or a “template”, that enables systematic evaluation of the effects of gait-training strategies on gait performance outcomes. This template can capture dominant features of bipedal locomotion without delving into the fine details of human anatomic structure and morphology. Key metrics of human walking gaits including the evolution of center of mass, joint patterns, ground reaction force etc. was demonstrated. Additionally, the template allows custom control schemes and characterization of joint torques for swing motion, driven by interlimb coordination in particular. This enables differentiating the performance of coordination strategy against the incumbent in gait stability, energetic efficiency, and response to speed modulations.

Through careful analysis of the model simulations and comparable experimental data, this proposed model not only proved its value in unifying the design and evaluation of gait-training strategies, also provided insights into the potency of a coordinationcentric approach as a gait-training strategy.