Position Control of Flexible Joint Robots with PD Controller Based on Disturbance Observer

With the trend of “human-robots interaction”, robots will inevitably come into contact with people in task execution. In order to ensure human security, flexible reduction elements are actively introduced into robot joints. Compared to traditional rigid robots, Flexible Joint Robots (FJRs) have advantages on these aspects, such as low output impedance, energy saving and shock tolerance. However, their position control problem still faces great difficulties, mainly because of uncertainties in the actuation and external environment. In this paper we propose an feasible method, which integrates a disturbance observer (DOB) and a proportional-derivative (PD) controller. The lumped disturbance, including modeling inaccuracy and external uncertainty, is estimated by the DOB, while PD controller is utilized to stabilize the system. It is mathematically proven that the global asymptotic convergence of closed-loop system is guaranteed. The effectiveness of the proposed method will be verified on a self-built 2 degrees of freedom (DOF) planar FJR platform and experimental results demonstrate the good performance of this approach.