Impedance Control of Flexible Robot Manipulators

Abstract

Force control is very important for many practical manipulation tasks of robot manipulators, such as, assembly, grinding, deburring that associate with interaction between the end-effector of the robot and the environment. Force control of the robot manipulator can be break down into two categories: Position/Force Hybrid Control and Impedance Control. The former is focused on regulation of both position and manipulation force of the end-effector on the tangent and perpendicular directions of the contact surface. The later, however, is to realize desired dynamic characteristics of the robot in response to the interaction with the environment. The desired dynamic characteristics are usually prescribed with dynamic models of systems consisting of mass, spring, and dashpot. Literature demonstrates that various control methods in this area have been proposed during recent years, and the intensive research has mainly focused on rigid robot manipulators. On the research area of force control for flexible robots, however, only a few reports can be found. Regarding force control of the flexible robot, a main body of the research has concentrated on the position/force hybrid control. A force control law was presented based on a linearized motion equation[l]. A quasi-static force/position control method was proposed for a two-link planar flexible robot [2]. For serial connected flexible-macro and rigid-micro robot arm system, a position/force hybrid control method was proposed [3]. An adaptive hybrid force/position controller was resented for automated deburring[4]. Two-time scale position/force controllers were proposed using perturbation techniques[5],[6]. Furthermore, a hybrid position-force control method for two cooperated flexible robots was also presented [7]. The issue on impedance control of the flexible robot attracts many attentions as well. But very few research results have been reported. An impedance control scheme was presented for micro-macro flexible robot manipulators [8]. In this method, the controllers of the micro arm (rigid arm) and macro arm (flexible arm) are designed separately, and the micro arm is controlled such that the desired local impedance characteristics of end-effector are achieved. An adaptive impedance control method was proposed for n-link flexible robot manipulators in the presence of parametric uncertainties in the dynamics, and the effectiveness was confirmed by simulation results using a 2-link flexible robot [9].