Meeting Abstract
Many species of bipedal runners, such as cursorial birds, can run at a variety of speeds. However, each species has a pattern for choosing gait features (e.g. stride length – SL, stride frequency – SF, and duty factor – DF) for achieving any selected speed. This work uses theoretical math models combined with optimal control methods to predict these gait features across speeds by minimizing energy cost. Specifically, this work compares a spring-legged math model with swing costs against the experimental gaits of helmeted guinea fowl (Numida meleagris) during steady running across speeds. A three-parameter fit (spring stiffness, damping constant, and leg inertia) generated steady gaits on rigid terrain from 1.3m/s to 3.1 m/s with SL, SF, and DF similar to measured guinea fowl data – all as a consequence of energy minimization. These parameters are fitted once for the species, and are constant across speeds and terrain conditions. Further, modeling the terrain as a dissipative surface (e.g. sand or soft soil) predicts an increased DF, consistent with experimental data. We are currently testing the broader ability for the model to predict gait features of species with varied leg length and inertia relative to body mass (e.g. red-legged seriema (Cariama cristata) and elegant crested tinamou (Eudromia elegans)). Preliminary analysis suggests that birds with larger leg inertia prolong their flight phases as the model predicts. In ongoing work, we are applying this modeling framework to multi-step manuevers, such as a 90-degree turns, to test scenarios that require higher-level decision-making.
