Evolutionary swarm robotics: genetic diversity, task-allocation and task-switching

Tuci, Elio (2014) Evolutionary swarm robotics: genetic diversity, task-allocation and task-switching. Swarm Intelligence. In: ANTS 2014: International Conference on Swarm Intelligence, 10-12 Sept 2014, Brussels, Belgium. ISBN 9783319099514. ISSN 0302-9743 [Conference or Workshop Item] (doi:10.1007/978-3-319-09952-1_9)

Abstract

The goal of this study is to investigate the role of genetic diversity for engineering more resilient evolutionary swarm robotic systems. The resilience of the swarm is evaluated with respect to the capability of the system to re-distribute agents to tasks in response to changes in operating conditions. We compare the performances of two evolutionary approaches: the clonal approach in which the teams are genetically homogeneous, and the aclonal approach in which the teams are genetically heterogeneous. We show that the aclonal approach outperforms the clonal approach for the design of robot teams engaged in two task-allocation scenarios, and that heterogeneous teams tend to rely on less plastic strategies. The significance of this study for evolutionary swarm robotics is discussed and directions for future work are indicated.

Item Type: Conference or Workshop Item (Paper)
Additional Information: Cite this paper as:
Tuci E. (2014) Evolutionary Swarm Robotics: Genetic Diversity, Task-Allocation and Task-Switching. In: Dorigo M. et al. (eds) Swarm Intelligence. ANTS 2014. Lecture Notes in Computer Science, vol 8667. Springer, Cham
Research Areas: A. > School of Science and Technology > Computer Science > Artificial Intelligence group
Item ID: 21944
Depositing User: Elio Tuci
Date Deposited: 13 Jun 2017 14:19
Last Modified: 26 Sep 2018 13:54
URI: https://eprints.mdx.ac.uk/id/eprint/21944

Actions (login required)

View Item View Item