Nonlinear Motion Control of Humanoid Robot Upper-Body for Manipulation Task

Abstract

This paper presents nonlinear control algorithm for motion control of humanoid robot upper-body. Upper-body consists of two arms, each having seven degrees of freedom (DOFs), and multi-segment lumbar spine with six DOFs which enables motion of the trunk, increases the workspace of robot arms and contributes to anthropomorphic appearance of the robot movements. Manipulation task, where robot is supposed to move an object of unknown mass, in presence of parameter uncertainties and external disturbance has been considered. Weight of the object has been considered as an external disturbance. Nonlinearity of the robot dynamical model and coupling between robot segments have been taken into account during control design. Sliding mode control with disturbance estimator has been used in order to provide accurate trajectory tracking in presence of disturbances. Efficiency of the proposed control algorithm is verified through a numerical simulation and results are presented.