Optimizing Myocontrol for Daily-Life Activities

Kurzfassung

Communication between humans and machines is rapidly improving through advancements in speech-processing and developments like a touchscreen. Human-Machine-Interfaces are especially crucial in the field of self-powered prostheses. An ideal prosthesis would provide the user with a natural control like experience. Unfortunately, this goal is still not achieved. Especially self-powered hand prostheses are struggling with the simultaneous proportional control of multiple degrees of freedom (DOF). Even just adding the option of a wrist rotation leads to instabilities of the control system. Therefore, the goal of this thesis is, to improve myocontrol, by analyzing the so-called interference problem in surface electromyography-based prosthetic control systems. Additionally, it is intended, to find a metric that identifies problematic actions or even action combinations.