Analysis and planning of optimal and collision-free paths for industrial robots with a prismatic joint (robotics, path planning)
C. Campbell · Purdue e-Pubs (Purdue University System) · 1985
This thesis deals with automating the path planning process for Stanford manipulators. Obstacles are modelled by wire-frames, and the robot by a line segment. Obstacle expansion is used as compensation to achieve the latter simplification. Pseudo-obstacles are introduced which delineate unsafe positions for the end effector, thereby enabling the robot to be modelled as a point for path planning purposes. Safe trajectories are then planned, and various heuristic rules are then presented which modify the trajectory into a locally optimized path in the mean-squared distance sense. Algorithms which determine the safety of straight line segment paths are also presented as a necessary part of the path planning process.