Control Systems Approach to Balance Stabilization during Human Standing and Walking.

postural control;balance models;biomechanics;Mechanical Engineering;Engineering;Mechanical Engineering

Wu, Amy R.Gillespie, Brent ;

University of Michigan

关键词: postural control;    balance models;    biomechanics;    Mechanical Engineering;    Engineering;    Mechanical Engineering;   

Others  :  https://deepblue.lib.umich.edu/bitstream/handle/2027.42/116654/amyrwu_1.pdf?sequence=1&isAllowed=y

瑞士|英语

来源: The Illinois Digital Environment for Access to Learning and Scholarship

【 摘 要 】

Humans rely on cooperation from multiple sensorimotor processes to navigate a complex world. Poor function of one or more components can lead to reduced mobility or increased risk of falls, particularly with age. At present, quantification and characterization of poor postural control typically focus on single sensors rather than the ensemble and lack methods to consider the overall function of sensors, body dynamics, and actuators. To address this gap, I propose a controls framework based on simple mechanistic models to characterize and understand normative postural behavior. The models employ a minimal set of components that typify human behavior and make quantitative predictions to be tested against human data.This framework is applied to four topics relevant to daily living: sensory integration for standing balance, limb coordination for one-legged balance, momentum usage in sit-to-stand maneuvers, and the energetic trade-offs of foot-to-ground clearance while walking. First, I demonstrate that integration of information from multiple physiological sensors can be modeled by an optimal state estimator. I show how such a model can predict human responses to conflict between visual, vestibular, and other sensors and use visual perturbation experiments to test this model. Second, I demonstrate that feedback control can model multi-limb coordination strategies during one-legged balance. I empirically identify a control law from human subjects and investigate how reducing stance ankle function necessitates greater gains from other limbs. Third, I show the advantages of momentum usage in sit-to-stand maneuvers. Counter to many human movements, this strategy is not performed with energetic economy, requiring excess mechanical work. However, with optimization models, I demonstrate that momentum serves to balance effort between knee and hip. Fourth, I propose a cost model for preferred ground clearance during swing phase of walking. Walking with greater foot lift is costly, but inadvertent ground contact is also costly. Therefore the tradeoff between these costly measures, modulated by movement variability, can explain expected cost of ground clearance. These controls-based models demonstrate the mechanisms behind normative behavior and enables predictions under novel situations. Thus these models may serve as diagnostic tools to identify poor postural control or aid design of intervention procedures.