Powered Lower Limb Prostheses

摘要

Human upright locomotion emerged about 6 million years ago. It is achieved by a complex interaction of the biological infrastructure and the neural control. Bones, muscles, tendons, central nervous commands and reflex mechanisms interact to provide robust and efficient bipedal movement patterns like walking or running. Next to these locomotion tasks humans can also perform complex movements like climbing, dancing or jumping. Diseases or traumatic events may cause the loss of parts of the biological infrastructure or the ability to control the lower limbs. Thus an identification of the required framework helps to improve on the artificial lower limb design and the control for bipedal robots, exoskeletons, orthoses or prostheses. A first artificial leg design was reported about 5000 years ago. After losing one leg in a battle an iron leg was fitted to Queen Vishpla to get her back on the battlefield. Since this time major changes in the structure, the material and the functionality led to improved prosthetic restoration of physically disabled. The characteristics of the biological leg structure are imitated by technical components. Using carbon fiber for the design of prosthetic feet made it possible to benefit from the elastic recoil like in the Achilles tendon in stance phase. Dampers in prosthetic knee joints are able to mimic eccentric muscle work during the gait cycle. Clutch-like mechanisms are used to lock the knee during stance. Such a function is comparable to isometric muscle work. Semiactive knee joints allow changes in damping ratio to adapt the mechanical joint properties to the requirements. Using integrated force or inertial sensors, movement tasks can be identified. An adaptation of damping to different walking speeds and conditions, such as walking inclines, declines, or climbing stairs is possible. All these developments permitted that amputees gait got closer to the natural human gait pattern. However, until the end of the 20th century prostheses were not able to reproduce concentric muscle work. External positive energy is required to compensate for energy losses during locomotion. For climbing stairs or walking inclines not only the ankle, but also the knee joint contributes net positive work to lift the body center of mass. To achieve desired joint motion, a power source like a motor would be required that can inject energy to mimic the concentric function of the muscle fascicles. The thesis comprises an analysis of joint requirements, it evaluates the current prosthetic design approaches and develops models on artificial muscles to mimic lower limb biomechanics in walking and running. The developed models are biologically inspired, while motors represent the function of muscle fibers and springs represent the function of the tendons. These systems are optimized for criteria like minimum joint peak power or minimum required energy for the power source (motor). Results demonstrate that elastic elements can highly decrease the actuator requirements. The springs are able to store energy in one phase of the gait cycle and to release it later when high peak power is required. Without the elastic assistance the reproduction of human joint behavior is hardly possible using current motor technology. The optimized interaction of motor and elasticity is evaluated in walking and running, using a prototype of a powered ankle prosthesis (Walk-Run ankle, Springactive). Next to experiments with a nonamputee, where the prosthesis was fitted in parallel to the fixed healthy ankle joint (Bypass), also experiments with a female unilateral transtibial amputee were performed. The optimized model behavior was compared to experimental observations and showed good agreement. Furthrmore, a concept on the improvement of an optimized walking motor pattern was successfully tested. By smoothening the motor curve to the main characteristics (low-pass filter) it was possible to increase the mechanical work output, to improve the system e