| A human ambulator continuously adapts to changing terrain conditions by modifying the dynamics of movement of its own components and their relationship to the external world. The study of human locomotion as a biological model of adaptation, inspires research in a variety of disciplines.; In this work, the study of adaptation of human gait is carried out by the development of neuro-musculo-skeletal models. The models developed proceed from the skeletal system inwards; i.e., the skeletal system is modelled first, the muscles are added next, and finally the central control system is developed.; Skeletal models (rigid bodies) are developed with various levels of complexity (2-D and 3-D). The results show translational energy applied at the hip joint provides a given toe elevation for minimal energy cost. This strategy is most effective when initiated during the double support phase, and results in increased hip elevation velocity at toe-off.; Next, bi-articular muscles are added, and the muscle force inputs are optimized to best satisfy the postulated objectives for landing stability, obstacle clearance, and efficiency of movement. The simulation results demonstrate that the use of bi-articular muscles is sufficient to clear a range of obstacles with the trailing limb (obstacle encountered during early swing). Stride length or landing stability objectives need not be specified, suggesting a simpler control of trailing limb trajectory by the central nervous system (CNS).; A proactive (feed-forward) control system for a novice ambulator is developed by combining a neural network model, fuzzy logic control, genetic optimization, and reinforcement learning. The ambulator, after performing several trials, learns to relate visual inputs of the obstacle size and location to limb movement dynamics.; The neuro-fuzzy-genetic model is also used for performing muscle force optimization. Some of the findings are: (a) for the leading limb stepping over an obstacle, less active control during the swing phase and more planning during the double support phase are required; (b) for the trailing limb, modifying the initial velocities at toe-off is not sufficient to achieve obstacle clearance and landing stability, and the addition of muscles is required to produce a satisfactory trajectory; and (c) the model is able to achieve an adaptive behavior.; Finally, the modelling approach is applied to gain insight to reactive balance control strategies during locomotion. Specifically, the available response time for recovery from an unsuccessful step over an obstacle (tripping) is considered. The simulation results show that the available response time increases with obstacle compliance, size, and its location within the stride. (Abstract shortened by UMI.)... |
