Motion planning and dynamic control of a linked manipulator using modified magnetic fields

Abstract

This work targets the navigation and control of multi-linked manipulators in tight, constrained spaces as are encountered in problems of robotic packaging or loading and unloading of pallets. The modified magnetic field algorithm was introduced by the authors (1996, 1997) for the specific purposes of generating motion plans for a robot moving in a complex and constrained workspace while veering around enclosing Cartesian space obstacle walls. The modified magnetic field method allows global convergence and collision constraint satisfaction while allowing the application of this method as a dynamic online plan generation and control technique in both Cartesian and configuration spaces. Methods for using the MMF for the online dynamic closed loop control of a robot arm in the presence of Cartesian obstacles and singular configurations are developed in this paper and results shown for a specific workspace configuration. This is done by extending the concept of the MMF through a detailed analysis of exactly what constitutes a desirable path in both Cartesian and configuration spaces. The dynamic control is performed in a hybrid manner to account for both, Cartesian as well as configuration space obstacles.