Reinforcement Learning of Manipulation and Grasping Using Dynamical Movement Primitives for a Humanoidlike Mobile Manipulator

It is important for humanoid-like mobile robots to learn the complex motion sequences in human-robot environment such that the robots can adapt such motions. This paper describes a reinforcement learning (RL) strategy for manipulation and grasping of a mobile manipulator, which reduces the complexity of the visual feedback and handle varying manipulation dynamics and uncertain external perturbations. Two hierarchies plannings have been considered in the proposed strategy: 1) high-level online redundancy resolution based on the neural-dynamic optimization algorithm in operational space; and 2) low-level RL in joint space. At this level, the dynamic movement primitives have been considered to model and learn the joint trajectories, and then the RL is employed to learn the trajectories with uncertainties. Experimental results on the developed humanoidlike mobile robot demonstrate that the presented approach can suppress the uncertain external perturbations.