Measuring the angle of a rotating link through compliant driving tendons

Measuring the position of a tendon-driven rotating link, based on the state of the tendons at the proximal end is challenging. Specifically, tendon elasticity and Coulomb friction can cause significant nonlinearities in the relationship between the motion at the proximal end of the tendon and the motion of the link. An extended Kalman filter (EKF) is proposed as a potential method to refine estimates of link position based on proximal measurements. A model link and tendon system is built to demonstrate the impact of these effects on measurement. Sensors are incorporated to allow direct measurement of the link position in addition to the motion of the tendons at their origins. A dynamic model is developed for the resulting system and used in conjunction with an EKF to generate improved estimates of the linkpsilas motion based on proximal measurements alone.