Skip to main content
Fig. 1 | Journal of NeuroEngineering and Rehabilitation

Fig. 1

From: Human arm weight compensation in rehabilitation robotics: efficacy of three distinct methods

Fig. 1

The three arm weight compensation methods: Method Average: Masses \(\tilde {m}\) and center of masses (\(\tilde {CoM}\)) are approximated through anthropometric tables based on basic information of the subject. The arm is modelled as two independent rigid bodies, upper arm (ua) and lower arm (la). The CoMs of the objects are assumed to lie on the arm segment rotation axis. Method Full: The parameters \(\hat {m}\) and \(\hat {CoM}\) are estimated through force/torque sensor data at each cuff. The arm is modeled as two independent rigid bodies, similarly to method Average, though the \(\hat {CoM}s\) of each object are not assumed to lie on arm segment rotation axis and are fully estimated. Method Equilibrium: A quasistatic equilibrium against gravity is created at each cuff. Upper and lower arm are assumed to be connected to each other. The passive joint torques τp,ua and τp,la are partially taken into account. The support force parameters \(\hat {f}_{ua}\) and \(\hat {f}_{la}\) are calculated and achieve an equilibrium against the wrenches created by gravity at the cuffs, ωmeas,ua,ωmeas,la. The CoMs of the arm are assumed to lie on each arm segment rotation axis

Back to article page