In this paper, we address the problem of path planning for a revolute manipulator, operating in a workspace filled with obstacles whose boundaries are enveloped by circle or possibly any other shape. The path planning approach presented here is developed in the manipulator joint space and consists of two strategies. The first strategy is used in tightly packed free spaces located between the workspace obstacles, this approach relies on constrained optimization. The second strategy is used whenever the manipulator is operating in free space regions which are not surrounded by the workspace obstacles, this strategy does not use optimization but relies on a global but safe approximation of the free space regions of the joint space. The manipulator uses the second strategy to perform collision-free gross motions. This division of the path planning approach reduces the chances that the manipulator will get stuck in a geometrical local minimum if the constrained optimization procedure is used to entirely determine the path planning procedure. With the approach taken here, the chance of finding a path for the manipulator from its given start point to its desired goal point in an automated manner is high in comparison to one which uses just one of the above two strategies.

Date of this Version

March 1993