Coordenação ótima de múltiplos robôs de serviço e de recarga em tarefas persistentes

Abstract

Autonomous multi-robots can be used to perform tasks that require cooperation between them. This work consists at the coordination problem of land mobile robots, divided into two groups: working robots and recharge robots. The working multi-robots must travel by closed paths that intersect each other. A high-level controller acts directly on average speeds of these vehicles, avoiding possible collisions and ensuring security in their tasks. The paths are traversed in cycles, which should be commensurate and persistents, executed on a infinite planning horizon, which overcomes the limitations of charges your batteries. For this is added one dedicated group of robots that operates in the battery exchange process. This group is called recharge robots. The strategy used to recharge all working robots is based on graph. The coordination of multiples robots is resolved through two optimizers, both implemented with LINGO solver, integrated with ROS (Robot Operating System) using the C++ language. One of optmizers coordinates the movement of working robots in order to avoid collisions between them. The results obtained in this first optimization step are used for the two robotic groups, became they must remain in line during the charging process. Thus, the second part of the optimization is to determine the optimum paths to be traced by the recharge robots. This problem is solved by a mixed integer linear program (MILP), with the objective of minimizing the overall time for recharging task.