German Conference on Robotics · 2010 · 14 citations · 0 references
EngineeringMechanical EngineeringRobotics EngineeringMotion PlanningPath PlanningRoboticsCell DecompositionCollision DetectionRapidly Exploring Random TreesRehabilitation RoboticsRobot ManipulatorsRobot Motion PlanningBi-directional Rapidly Exploring Random TreesFree Configuration SpaceMotion Planning For ManipulatorsMotion Planning With Rapidly Exploring Random TreesRandom TreesRobot Path PlanningRobotic Path PlanningTrajectory Planning
This paper presents a new approach called Cell based Bi-directionall Rapidly exploring Random Trees (CellBiRRT) for manipulator motion that is able to handle pose constraints and to operate in difficult environments with high probability of collisions. The approach uses the idea of bi-directional RRT (Bi-RRT) and creates an algorithm that uses configuration (C-Space) and Cartesian space. It is achieved by decomposing the Cartesian space into cells, which will be used later as basis in order to reduce the free C-Space by generating N-dimensional cuboid regions. The results show that the method can be an acceptable solution to path planning problem, allowing a 7 Degrees of Freedom (DoF) robot arm to perform household?s activities, like moving and placing bottles or keeping horizontally a meal tray in order to serve handicapped people.