A new approach for robot Path Planning with Genetic Algorithms

Abstract

Path planning has important applications in many areas, for example industrial robotics, autonomous systems, virtual prototyping, and computer-aided drug design. This paper examines robot path planning. This is the task of planning the motionsof a robot so that it avoids collisions with objects in the workspace.The method is applied to a twolink planar robot manipulator using a two dimensional conguration Space representation Parallelism within the method is identified and reductions in execution time are achieved. The basic concept of the Tangent graph is extended to develop a new path planner. The second new planner is developed based around a Genetic Algorithm. Thisprovides a probabilistic approach to the search for a path. This is also demonstrated using a simulation of a three link planar manipulator andagain parallelization issues are discussed. The implications of incorporating the new path planners into a robot system areconsidered. A practical testing is described which integrates a vision systempath planner trajectory planner and manipulator controller.