%0 Generic
%A Stein, Kevin
%C Heidelberg
%D 2021
%F heidok:30924
%R 10.11588/heidok.00030924
%T Experimental and Computational Stability Analysis:  from Slackline Athletes to Persons with Schizophrenia
%U https://archiv.ub.uni-heidelberg.de/volltextserver/30924/
%X Humans have walked the earth for more than 200,000 years, yet it is not fully understood how this is  achieved in stable and robust manner. In this work we analyzed human balancing during static and  dynamic balance tasks based on motion capture data. A balancing task requires the subject to either  constantly perform recovery movements or to learn an inherently stable and robust motion specific to  the task. We designed a static balance test and analyzed dynamic balance during slackline balancing  and the tandem walk test. We gathered data from over 60 participants in two studies.  In the first study, we analyzed slackline balancing with the goal to define balance performance indicators  and measures for slackline expertise. We compared beginners that had never balanced on a slackline  before to professional slackline athletes. All participants also performed the static balance test. We  found that trained slackliners balance very well in the static balance test, whereas the beginner group  showed a larger variance in the time they managed to balance. Therefore, we divided the beginner  group into balance-experienced and balance-inexperienced according to this test. Based on over 300  balancing trials on the slackline of 20 participants we defined and evaluated over 30 balance metrics.  Normalized angular momentum and Center of Mass acceleration allow us to quantify stability and  amount of recovery movements. Posture and movement was similar for the beginner groups, whereas  professional slackliners have adapted a different pose and strategy that allows them to consistently  maintain a horizontal head orientation and upright posture. We found that their hand movement is more  coordinated, their stance foot less accelerated by the slackline and their stance leg more compliant.  In the balance study of the "Schizophrenia and the Moving Body" project by Lily Martin, we compared  the balance capabilities of an experimental group consisting of persons with schizophrenia to a healthy  control group. Participants performed the static balance test and twice the tandem walk, once with eyes  open and once with eyes closed. Applying the balance indicators showed deficits in static and dynamic  balance of the experimental group when compared to the control group. They had significantely larger  values for normalized angular momentum and Center of Mass acceleration and took significantly more  recovery steps to maintain balance. When analyzing the strategy employed by the two groups, we found  that the control group successfully used their arms to balance and place correct steps, whereas the  experimental group mainly relied on recovery steps and did not involve the arms.  We combined and analyzed data from all participants of both studies. For static balance we found  that sideways Center of Pressure sway distance and front-back sway velocity are the most suitable  parameters to quantify balance capabilities. We then compared flat ground tandem walking to slackline  walking based on the balance performance indicators. Professional slackline athletes are able to reduce  many performance indicators to the values regular tandem walking. The only show larger normalized  angular momentum in the frontal plane and increased sideways Center of Mass acceleration, which are  both a direct consequence of the instability introduced by the slackline. Further, we found that their  pose and movement strategy is specific to slackline balancing and not used by other participants neither  in tandem walking nor slackline walking of beginners.  In the last part of this thesis we prototyped a pressure sensor for the slackline and showed that we  can reliably measure Center of Pressure data during slackline balancing. Based on the findings we  developed a specific contact interaction model. We used this contact model inside an optimal control  problem formulation to perform a fully dynamic reconstruction of slackline jumping