Operating a robotic network involves addressing feedback control, task planning and dynamic resource assignment. Ideally, these issues should be tackled in an integrated way, although this poses a number of complex challenges involving combinatorial complexity, computational requirements and implementation efforts. This paper presents a novel dynamic task assignment algorithm integrated in a discrete event formalism, which is able to fully characterize the dynamics of the network and incorporate robots allocation, conflict resolution and task assignment policies. After providing a brief introduction to the main characteristics of the network model and control framework, this paper focuses on the task scheduling algorithm. The effects of the various configuration parameters, along with the strengths and limitations of the proposed approach are assessed by means of an extensive simulation study considering robotic networks of various sizes and characteristics
A Heuristic Approach to Task Assignment and Control for Robotic Networks / Di Paola, D.; Naso, David; Turchiano, Biagio. - (2010), pp. 1784-1790. (Intervento presentato al convegno IEEE International Symposium on Industrial Electronics, ISIE 2010 tenutosi a Bari, Italy nel July 4-7, 2010) [10.1109/ISIE.2010.5637599].
A Heuristic Approach to Task Assignment and Control for Robotic Networks
NASO, David;TURCHIANO, Biagio
2010-01-01
Abstract
Operating a robotic network involves addressing feedback control, task planning and dynamic resource assignment. Ideally, these issues should be tackled in an integrated way, although this poses a number of complex challenges involving combinatorial complexity, computational requirements and implementation efforts. This paper presents a novel dynamic task assignment algorithm integrated in a discrete event formalism, which is able to fully characterize the dynamics of the network and incorporate robots allocation, conflict resolution and task assignment policies. After providing a brief introduction to the main characteristics of the network model and control framework, this paper focuses on the task scheduling algorithm. The effects of the various configuration parameters, along with the strengths and limitations of the proposed approach are assessed by means of an extensive simulation study considering robotic networks of various sizes and characteristicsI documenti in IRIS sono protetti da copyright e tutti i diritti sono riservati, salvo diversa indicazione.
