Geometric Path Planning for General Robot Manipulators
Ziyad Aljarboua · 2009
In this paper, we present a distance transform-based geometric path planning algorithm suitable for robots with vision capability. We show that the proposed approach can reduce the computation time needed to find an optimal collision-free path compared to other path planning algorithms by utilizing the output of the computer vision module and by eliminating any extra work to model the workspace of the robot. We illustrate the application of this algorithm in 2D and 3D and present two optimization algorithms to reduce the number of waypoints and shorten the Euclidean distance of the path. Finally, we show how this algorithm can be incorporated in systems with on-board sensors and how the output of the vision module can be streamed to the algorithm to allow real-time path planning. Geometric path planning is a robust alternative algorithm for computing a collision free path connecting the initial and final configurations of a robot with a minimal number of waypoints in 2D and 3D environments. It provides a geometric description of the robot motion given a mapping and a description of the obstacles in the workspace. The algorithm outputs the (x,y,z) coordinates of the path in the workspace which is then passed to a custom inverse kinematics block to compute the revolute and prismatic joint variables in the configuration space. In this paper, we illustrate the application of this algorithm by computing an optimal path for the end-effector; however, this algorithm can be applied to all the origins of the Denavit Hartenburg frames or a set of floating control points in the links to ensure that the entire kinematic chain follows a collision free path. The motivation for this path planning approach is to minimize the computation time needed to find a collision free path for sensor-equipped robots by eliminating the representational cost incurred by modeling the workspace. This is done by designing the path planning algorithm around the output of the vision module, mainly the segmented image of the workspace.