Task-space dynamics and motion/force control of fixed-base manipulators under reaction null-space-based redundancy resolution

Abstract

SUMMARYThis paper introduces a task-space dynamics formulation for fixed-base serial-link kinematically redundant manipulators and a motion/force controller based on it. The aim is to alleviate joint-space instability problems that have been observed with other motion/force controllers. The dynamics are represented in floating-base coordinates, wherein the end effector is regarded as the floating base. This representation gives rise to a momentum-conserving redundancy resolution scheme based on the reaction null-space (RNS) method used in past studies on free-floating and flexible-base space robots. A generalized inverse is obtained that is shown to satisfy the conditions for dynamic consistency in the sense of the operational space (OS) formulation, but may lead to the joint-space instabilities observed earlier. The proposed controller is based on the pseudoinverse of the coupling inertia matrix and ensures reactionless link motion that does not disturb the force balance at the end effector. The performance of the RNS motion/force controller is examined by comparison to that with an OS motion/force controller. It is shown that while the performance in task-space of both controllers is satisfactory, the joint-space performance of the proposed controller is superior.

Research on coupling dynamics and coordinated control of a legged robot

“…The type of robots mentioned above has a distinct feature, that is, these robots exhibit high dynamical coupling between an actively operated part (manipulator) and a passively or actively moving part (a legged robot in this text). This coupling dynamics (Huang et al, 2015;Nenchev et al, 2016) is termed as reaction dynamics. Reaction dynamics can be used to deal with problems such as trajectory tracking or attitude adjustments of a free-floating robot (Lampariello et al, 2018;Misra and Bai 2017).…”

Active isotropic compliance in redundant manipulators

“…In fact, redundancy can be used to meet user-defined tasks or additional constraints on the kinematic or dynamic control problem, such as obstacles [22,23,24] or joint-drift [25] avoidance, joint limitations [26,27,28], singularity avoidance [29,30], energy saving optimization [31,32,33], torque [34,35,36,37] and antagonistic stiffness [38] optimization, force or torque control [39,40,41,42], impact force reduction [43], collision safety evaluation [44], to name the most relevant ones.…”

Bi-criteria minimization with MWVN–INAM type for motion planning and control of redundant robot manipulators

“…[1][2][3][4][5] Various approaches to the redundancy resolution of robot manipulators have been developed and studied. [1][2][3][6][7][8][9][10][11][12][13] The minimum two-norm approach, which minimizes the sum of the squares of joint variables (e.g., joint velocities, joint accelerations, and joint torques), is a common solution to the redundancy-resolution problem. 1,6,7,[13][14][15][16][17] The in-depth research on robotic redundancy resolution also yields a common variation of this minimum two-norm solution, termed weighted solution, which minimizes a weighted two-norm.…”