Forner-Cordero Arturo, Ackermann Marko, de Lima Freitas Mateus
Biomechatronics Laboratory, Mechatronics and Mechanical Systems Department, EscolaPolitécnica, University of São Paulo, São Paulo, Brazil.
Annu Int Conf IEEE Eng Med Biol Soc. 2011;2011:7829-32. doi: 10.1109/IEMBS.2011.6091929.
Perturbations during human gait such as a trip or a slip can result in a fall, especially among frail populations such as the elderly. In order to recover from a trip or a stumble during gait, humans perform different types of recovery strategies. It is very useful to uncover the mechanisms of the recovery to improve training methods for populations at risk of falling. Moreover, human recovery strategies could be applied to implement controllers for bipedal robot walker, as an application of biomimetic design. A biomechanical model of the response to a trip during gait might uncover the control mechanisms underlying the different recovery strategies and the adaptation of the responses found during the execution of successive perturbation trials. This paper introduces a model of stumble in the multibody system framework. This model is used to assess different feedforward strategies to recover from a trip. First of all, normal gait patterns for the musculoskeletal system model are obtained by solving an optimal control problem. Secondly, the reference gait is perturbed by the application of forces on the swinging foot in different ways: as an instantaneous inelastic collision of the foot with an obstacle, as an impulsive horizontal force or using a force curve measured experimentally during gait perturbation experiments. The influence of the type of perturbation, the timing of the collision with respect to the gait cycle, as well as of the coefficient of restitution was investigated previously. Finally, in order to test the effects of different muscle excitation levels on the initial phases of the recovery response, several muscle excitations were added to selected muscles of the legs, thus providing a simulation of the recovery reactions. These results pave the way for future analysis and modeling of the control mechanisms of gait.
人类步态中的干扰，如绊倒或滑倒，可能导致跌倒，尤其是在老年人等脆弱人群中。为了在步态中从绊倒或 stumble 中恢复，人类会执行不同类型的恢复策略。揭示恢复机制对于改进针对跌倒风险人群的训练方法非常有用。此外，人类恢复策略可应用于为双足机器人步行器实现控制器，作为仿生设计的应用。步态中对绊倒反应的生物力学模型可能揭示不同恢复策略背后的控制机制以及在连续扰动试验执行过程中发现的反应适应性。本文介绍了多体系统框架中的 stumble 模型。该模型用于评估从绊倒中恢复的不同前馈策略。首先，通过解决最优控制问题获得肌肉骨骼系统模型的正常步态模式。其次，通过以不同方式对摆动脚施加力来扰动参考步态：作为脚与障碍物的瞬时非弹性碰撞、作为脉冲水平力或使用在步态扰动实验期间实验测量的力曲线。先前已经研究了扰动类型、与步态周期相关的碰撞时间以及恢复系数的影响。最后，为了测试不同肌肉兴奋水平对恢复反应初始阶段的影响，向腿部选定肌肉添加了几种肌肉兴奋，从而提供了恢复反应的模拟。这些结果为未来步态控制机制的分析和建模铺平了道路。
