Co-Evolutionary Algorithm for Motion Planning of Two Industrial Robots with Overlapping Workspaces. (21st January 2013)

• Record Type: Journal Article
• Authors: Curkovic, Petar; Jerbic, Bojan; Stipancic, Tomislav
• Is Part Of: International journal of advanced robotic systems. Volume 10:Number 1(2013)
• Publication Date: 2013-01-21
• Subjects: Evolutionary Algorithms -- Motion Planning -- Multi Robot Systems -- Optimization; Robotics -- Periodicals; Robotics; Periodicals; 629.892
• DOI: 10.5772/54991
• Languages: English
• ISSNs: 1729-8806

Abstract

A high level of autonomy is a prerequisite for achieving robotic presence in a broad spectrum of work environments. If there is more than one robot in a given environment and the workspaces of robots are shared, then the robots present a dynamic obstacle to each other, which is a potentially dangerous situation. This paper deals with the problem of motion planning for two six-degrees-of-freedom (DOF) industrial robots whose workspaces overlap. The planning is based on a novel hall of fame - Pareto-based co-evolutionary algorithm. The modification of the algorithm is directed towards speeding-up co-evolution, to achieve real-time implementation in an industrial robotic system composed of two FANUC LrMate 200 i C robots. The results of the simulation and implementation show the great potential of the method in terms of convergence, robustness and time.