In this work the problem of the coordination in industrial multi-robot systems is analyzed, that is to say the coordination of several robots that have to perform different tasks in the same workspace. A methodology is presented, for off-line coordination of several robots interacting in a 3D environment. In particular, parts of the 3D space in which the robots could collide are found out (mutex zones) and automata models for the robots tasks and for the zones created are built. Then, the tasks for each robot are scheduled in order to minimize the time of execution of all the tasks, preventing collisions and blocking situations. Since the complexity of the problem is very large, automatic procedures are needed. This is done with the help of a software named RobotStudio, provided by ABB Robotics, for three dimensional simulations, and with Supremica, a software program developed at the Department of Signals and Systems at Chalmers University of Technology in Gothenburg, for the analysis of automata and Petri nets. The main result of the work is the automatic synthesis of a discrete event controller that optimizes the total time of work and guarantees that the multi-robot system is collision free and non-blocking. Future work can regard optimal choice of zones of mutual exclusions and simulations in a parallel environment for the robots.
COORDINATION OF MULTI-ROBOT SYSTEMS USING SUPERVISORY CONTROL AND OPTIMIZATION
[Voti: 0 Media Voto: 0/5]