INVESTIGADORES
MAS Ignacio Agustin
congresos y reuniones científicas
Título:
Energy Efficient Trajectories for Coordinated Multi-Robot Systems
Autor/es:
IGNACIO MAS; KYLE STANHOUSE; CHRISTOPHER KITTS
Lugar:
Santa Clara
Reunión:
Congreso; 2011ASME Information Storage and Processing Systems Conference; 2011
Institución organizadora:
ASME
Resumen:
The coordinated motion of multiple mobile robots allows for the execution of tasks beyond the typical capabilities of single robots systems.  Redundancy, physical distribution of sensors or actuators and increased coverage are some of the advantages of deploying multi-robot systems. The real-time coordination of such complex systems is a topic of active interest in the robotics community. Once a multi-robot coordination control algorithm is implemented, choices remain to be made on how to operate the resulting system. In particular, energy cost is a consideration of vital importance when addressing the efficiency of task execution. In this article, a particular formation control approach called Cluster Space Control Framework is utilized to control a group of three non-holonomic wheeled robots. The approach allows for simple specification, control and monitoring of the state of the system at the application level, which is possible by the degree of abstraction introduced by the control framework. Based on this approach, this article addresses the problem of following a predefined time independent path in the application space by implementing alternative trajectories—which are functions of time—and analyzing the total energy cost of each implementation. This energy cost is calculated as the sum of the squared velocities of all the robots in the formation integrated over the duration on the trajectory. The results show that for a given fixed application space path, by varying the instantaneous velocities of the robots, the final configuration can be reached at the same final time while decreasing the total energy cost of the system. These results indicate the existence of an energy-optimal application-space trajectory, which can be found using common multi-variable optimization algorithms. This occurrence is a consequence of the variation over the path of the formation equivalent dynamics, such as the formation equivalent moment of inertia, which can be analytically derived from the robot dynamics and the pose of the formation at any given time.