Optimal motion planning of mobile manipulators with minimum number of platform movements

Abstract

In many practical applications the end effector of a mobile robotic manipulator is required to reach a set of desired positions and orientations. This is commonly achieved through a combination of platform repositionings and manipulator reconfigurations. Considering the fact that precise repositioning of the platform of an articulated mobile robot is complex and time-consuming, we propose a novel motion planning approach that would minimize the number of platform repositionings and manipulator reconfigurations for place to place tasks. The proposed approach would also guarantee that the joint limits are not crossed and obstacles are avoided. The approach comprises two steps. First the optimal manipulator configurations corresponding to desired end-effector poses are determined and then successive optimal configurations are connected with minimal paths. Rapidly-Exploring Random Trees (RRTs) are used to design configuration-connecting paths. Results of simulated experiments on a 10 DOF mobile manipulator have demonstrated the applicability and effectiveness of the proposed approach.