Mobile robot arm trajectory generation for operation in confined environments

Abstract

This article gives a novel trajectory generation algorithm which is suitable for operating in confined and dangerous environments. A technique is developed which can, in real time, calculate the existence of a safe path for navigating a robotic manipulator arm between obstacles without collision and accurately generating an efficient path between them. A map of the environment is created in the control servo domain using existing environment data where each dimension of the map space represents one of the degrees of freedom of the manipulator. The map is a multi-dimensional space that represents the control ranges of the manipulator and which contains obstacles to be avoided. The start and desired locations of the arm end effector can be converted into this space so that a path can be generated between them. The space is split into a graph of nodes, and a Dijkstra’s shortest path algorithm is used to generate a safe trajectory. If a successful trajectory can be found, then the arm desired location is achievable and the list of nodes that make up the trajectory form a set of control requirements that can be followed to drive the arm to its desired geometry.