Planificación de trayectorias en robots redundantes con mapas de factibilidad y RRT

Abstract

Redundant manipulators offer multiple advantages, such as improved manipulability, or evasion of singularities and obstacles. However, kinematic redundancy also introduces additional difficulties, such as the need to solve an underdetermined inverse kinematics problem. This article presents a novel method for motion planning of redundant manipulators, based on the exploration of feasibility maps. The proposed method is an extension of the Rapidly-exploring Random Trees (RRT) algorithm, modified to explore the redundant task space with the purpose of finding a feasible suboptimal path in the joint space, sacrificing optimality for scalability to a higher number of degrees of redundancy. The method is capable of following a given trajectory in the task space, while considering other constraints, such as obstacle avoidance or joint limits. The method has been validated under simulation