"Motion Analysis and Synthesis for Movement Assistance and Rehabilitation"
Motion Analysis and Synthesis for Movement Assistance and Rehabilitation
Sprache des Titels:
Over the last decades a continuous trend of the worldwide population growing older could be observed. Increasing life expectancy and falling birth rates are identified as the main factors driving this demographic change. The demand for motion assistance and rehabilitation of age-related movement disorders is growing accordingly, which also facilitates interest in robotic solutions. Two of such devices will be addressed in this thesis, a commercial robotic hippotherapy system and an exoskeleton for the human lower body, developed at the JKU Institute of Robotics. In connection with these robotic systems, several methods and approaches in the context of rehabilitation and motion assistance are introduced. A major topic thereby is the generation of artificial motion patterns based on motion capture data.
Hippotherapy refers to horse riding in the context of rehabilitation. It is a medical treatment that has been successfully employed in various fields, e.g. for improving locomotion performance of patients with movement disorders. Robotic systems enable the application of hippotherapy in clinical environments. Additional benefits, amongst others, are the continuously adjustable speed and high repeatability. Fundamental for a therapy outcome equivalent to classical hippotherapy is that the horse motion reproduced by the robotic system is as realistic as possible. Based on an analytical and time-continuous motion description a method to reproduce the horseback movement during typical horse gaits is presented in this thesis. This method allows for motion synthesis with any desired time span and time resolution, generating realistic trajectories applicable to robotic systems for riding simulation in general and robotic hippotherapy in particular. An adjustable time scaling parameter enables the adaptation of the generated motion according to the physical abilities of the patient or the capabilities of the robotic system.
Serving as testing platform, where novel control strategies and motion assistance approaches can be implemented in a low-level manner, an exoskeleton prototype was developed in the course of this thesis. In addition to the design and construction of the setup and the pre-requisite implementations facilitating communication and operability, robust recognition of the user's motion intention and accurate timing of the provided assistance are focused on. Two approaches addressing these issues, namely gait assistance by learning periodic motions, as well as classification of gait phases based on measured muscle activities, are introduced.
Physiologically consistent motion patterns of healthy human gait are intended to serve as a sound scientific basis for the developed methods and simulations. To this end, a holistic approach to simultaneously identify the geometric parameters of a kinematic human lower limb model and the parameters defining a cyclic gait trajectory, based on three-dimensional motion capture marker positions, is introduced. The presented methodology is then utilized to analyze several datasets of measured gait trials. The influence of person-specific factors, like age or physique, onto the motion pattern is analyzed by applying the presented biomechanical model to gait data of different male and female participants walking at various speeds. Modeling the dependency of gait motions on the considered variables enables the generation of artificial joint angle trajectories according to a simulated person with defined body parameters walking at a prescribed speed.
The presented methods show promising experimental results, which are deemed valuable for future research projects focusing on technologies for rehabilitation or movement assistance.