GA based Trajectory Planner for Robot Manipulators Sharing a Common Workspace

E.A. Merchán-Cruz and A.S. Morris (UK)


collision avoidance, genetic algorithms, multiple manipulators, trajectory planning.


This paper presents a Genetic Algorithm (GA) based trajectory planner for a system consisting of two robot manipulators. The systems considered in this paper are two 2 degrees of freedom (dof) planar manipulators and two 3 dof planar manipulators. Each manipulator is considered as a moving obstacle by the other and collision is avoided using the Potential Field approach. Both robots have the same hierarchy, and the proposed algorithm searches in parallel for adequate configurations without giving any special preference to either robot. The GA carries out optimisation in parallel to find the best set of configurations for both manipulators that causes no collisions, while minimising the error to their respective goals. The same fitness function is specified for both manipulators throughout the entire process.

Important Links:

Go Back