Shortest path planning for single manipulator in 2D environment of deformable objects

Abstract

A heuristic algorithm to perform path planning for single manipulator in 2D environment containing deformable objects is presented. The environment is partitioned into a quadtree hierarchy for both sampling and space navigation use before combination of artificial potential field and heuristic reasoning are applied iteratively to generate feasible path for the manipulator. The algorithm specifically targets for the shortest path without damaging any objects due to deep collision depth between manipulator link and object. Resulting path is in turn to be used in generating micro-instruction controlling the manipulator. Implementation results show feasibility to solve problems involving simple object and manipulator configuration.