Sampling based Constrained Motion Planning For Floating Base Manipulators Using Constraints Driven Alternative Parameterization

Constrained end-effector manipulation is in big demand in many robotic applications. We present a novel end-effector-constrained motion planning technique for floating-base manipulators. The proposed planner is superior to that cited in literature in terms of accuracy and execution time. The analytic comparison with the former approaches are given as well as simulation results for an end-effector trajectory following task. The strength of the proposed algorithm roots from the use of an alternative parametrization of the configuration space, where end-effector pose is included in this alternative parameterization space explicitly. This makes it possible to successfully generate constrained configuration samples with a uniform distribution. The proposed algorithm is based on RRT. We present a variation of the sample generation and steering routines which are suitable for end-effector constrained planning using the proposed alternative parameterization. Also, we present an asymptotically optimal version of our planner based on RRT* via simulations for end-effector trajectory tracking in presence of physical obstacles