The trajectory planning and path following control of an anthropomorphic upper limb

In this paper, we study the trajectory planning and the disturbance-based model primitive PID control of an anthropomorphic arm. Firstly, an improved Q learning algorithm is proposed to find the optimal trajectory of the end-point and the elbow joint. Secondly, an accurate path following control is used to achieve mild reaching movements of the anthropomorphic arm. A disturbance observer-based model predictive PID control is applied to make the anthropomorphic upper limb follow an accurate path. The path following control is verified by real experiments. Results show that the anthropomorphic arm may well emulate mild reaching movements by proposed methods.