Set-point navigation of a redundant robot in uncertain environments using finite range sensors

Abstract

Abstract:In this work, control of redundant robot manipulators in an uncertain environment is considered. The manipulator is equipped with finite range sensors to detect obstacles in its workspace. A navigation functionbased kinematic controller is proposed to ensure the regulation of the end-effector to a desired set-point while the entire manipulator simultaneously avoids the obstacle points detected by the sensors. A joint-space controller is then utilized to ensure asymptotic tracking of the desired jointspace trajectory.

A New Objective Function for Obstacle Avoidance by Redundant Service Robot Arms

“…Another previous work of the third author, in [27], a gradient-based Lyapunov-type version of the controller in [17] could have also been utilized. Finally, if there are unstructured uncertainties in the dynamics, then the robust controller in [15] can be utilized.…”

“…In view of the closed-loop null space velocity tracking error in (15), the auxiliary controller in (13) ensures that the joint velocities in the null space converge to ( I − J + J ) g, provided that the vector function g and its time derivative are designed to be bounded (the proof is provided in Appendix C).…”

“…In view of the closed-loop null space velocity tracking error in (15), the auxiliary controller in (13) ensures that the joint velocities in the null space converge to (…”

“…Alternatively, [14] preferred to utilize the reciprocal of the minimum distance as the objective function. Other studies, such as [15], equipped a redundant robot with multiple proximity sensors to avoid collision with obstacles. However, as highlighted in [16], utilizing the minimum distance in the objective function is problematic when there are multiple obstacles in the workspace of the robot.…”

“…Prior research addressed obstacle avoidance algorithm designs as a sub-task for redundant robots [10][11][12][13][14][15][16] and a review of null-space. In these studies, the common approach was to first identify the closest points on the links of the redundant robot to the obstacles and then design a sub-task objective function to keep those points away from obstacles.…”