Motion planning and redundancy resolution of a rover manipulator

Abstract

This paper proposes a cooperative motion planning and redundancy resolution technique for a rover manipulator working on rough terrain in a 3-dimensional workspace. The primary motion requirement is accomplished by the manipulator to cover a prescribed trajectory, whereas the rover moves to augment its workspace. The proposed approach involves formulating the trajectory planning as non-linear constrained optimization problem, in which the pose of the platform is computed to achieve the desired trajectory by satisfying joint angle constraints and manipulability measure of the manipulator. The method generates a natural optimized motion for the rover manipulator by solving the constrained optimization problem using the Augmented Lagrangian Method (ALM). Several simulations have been conducted and a number of test cases have been presented to establish the efficacy of the proposed method in providing acceptable path.

End-effector path tracking of a 14 DOF Rover Manipulator System in CG-Space framework

“…Collision avoidance requires an exterior penalty function in the optimization approach. Rekha et al [12] have applied a classical optimization approach named augmented Lagrangian method (AML) to resolve redundancy in rough terrain non-holonomic motion planning using the notion of lazy arm movement. Accuracy in the parametric analysis of robots is complex to obtain in real time.…”

“…Rekha et al [51] present a KSOM learning framework to solve the inverse kinematics of such a manipulator system moving on flat/smooth terrain and then uneven terrain later. They also use ALM based optimization method [12–52], which took 10–15 min to calculate the whole mobile manipulator pose corresponding to just a single point on the end-effector path. Rekha et al [53] modified the standard KSOM architecture to coordinate between the mobile base and arm while solving inverse kinematics with multiple kinematic constraints.…”

“…In most cases, hyper-redundant mobile manipulator systems have a simple mobile base as an omnidirectional robot moving on a flat terrain. However, there are a few research contributions [12–55] in which the redundancy resolution problem of the Rover Manipulator System has been addressed using optimization or learning-based methods. However, the optimization [12] based on redundancy resolution and planning approach could not achieve real-time implementations.…”

“…However, there are a few research contributions [12–55] in which the redundancy resolution problem of the Rover Manipulator System has been addressed using optimization or learning-based methods. However, the optimization [12] based on redundancy resolution and planning approach could not achieve real-time implementations. On the other hand, the learning-based approach needs extensive offline training procedures.…”

Simplified Motion Control of a Vehicle manipulator for the Coordinated Mobile Manipulation

“…Hence the main contribution of the paper is to propose as mentioned above, in this paper, a novel kinematic control scheme and its overall performance is compared to the conventional operational-space control scheme. To get an optimal solution, it is mandatory to employ some constrained optimisation techniques and few of them are approached in this manner and they are available in the literature [23][24][25][26][27][28] .…”