Evolving a Repertoire of Controllers for a Multi-function Swarm

Engebråten, Sondre Andreas; Moen, Jonas; Yakimenko, Oleg A.; Glette, Kyrre

doi:10.1007/978-3-319-77538-8_49

Cited by 16 publications

(25 citation statements)

References 21 publications

Supporting

Mentioning

Contrasting

Order By: Relevance

“…Going from two applications (exploration and networking) to three applications (exploration, networking and geolocation) this did not seem to be the case-rather the opposite. In a previous work a single repertoire could hold 608 solutions (Engebråten et al, 2018b), compared to the average of 2,031 solutions per repertoire in this work. The combination of an open-ended evolutionary method and a multi-function behavior repertoire seems to enable the evolution of repertoires that truly span the entire range of viable solutions.…”

Section: Discussionmentioning

confidence: 98%

“…Compared to previous works (Engebråten et al, 2018b), the ranges of the weight and scale parameters were reduced. For these experiments, the weight parameter is limited from −5.0 to 5.0 and the scaling parameter is limited from −0.5 to 0.5.…”

Section: Controller Frameworkmentioning

confidence: 88%

“…The contributions of this paper are an extensible and rigorous framework for multi-function swarms, incorporating automated behavior generation and methods for analyzing the resulting behaviors. This is a major extension of previous works (Engebråten et al, 2018b). Through the use of ablation, or the selective disabling or removal of parts of the controller, importance of individual sensory inputs is determined.…”

Section: Introductionmentioning

confidence: 86%

“…Three applications were chosen to evaluate the framework: area surveillance, communication network creation and geolocation of RF emitters. The first two were introduced in previous works (Engebråten et al, 2018b), while the combination with geolocation of RF emitters is new in this paper.…”

Section: A Tri-functional Swarmmentioning

confidence: 99%

See 3 more Smart Citations

A Framework for Automatic Behavior Generation in Multi-Function Swarms

et al. 2020

Self Cite

View full text Add to dashboard Cite

Multi-function swarms are swarms that solve multiple tasks at once. For example, a quadcopter swarm could be tasked with exploring an area of interest while simultaneously functioning as ad-hoc relays. With this type of multi-function comes the challenge of handling potentially conflicting requirements simultaneously. Using the Quality-Diversity algorithm MAP-elites in combination with a suitable controller structure, a framework for automatic behavior generation in multi-function swarms is proposed. The framework is tested on a scenario with three simultaneous tasks: exploration, communication network creation and geolocation of Radio Frequency (RF) emitters. A repertoire is evolved, consisting of a wide range of controllers, or behavior primitives, with different characteristics and trade-offs in the different tasks. This repertoire enables the swarm to online transition between behaviors featuring different trade-offs of applications depending on the situational requirements. Furthermore, the effect of noise on the behavior characteristics in MAP-elites is investigated. A moderate number of re-evaluations is found to increase the robustness while keeping the computational requirements relatively low. A few selected controllers are examined, and the dynamics of transitioning between these controllers are explored. Finally, the study investigates the importance of individual sensor or controller inputs. This is done through ablation, where individual inputs are disabled and their impact on the performance of the swarm controllers is assessed and analyzed.

show abstract

Section: Discussionmentioning

confidence: 98%

Section: Controller Frameworkmentioning

confidence: 88%

Section: Introductionmentioning

confidence: 86%

Section: A Tri-functional Swarmmentioning

confidence: 99%

See 2 more Smart Citations

A Framework for Automatic Behavior Generation in Multi-Function Swarms

et al. 2020

Self Cite

View full text Add to dashboard Cite

show abstract

“…Novelty-driven algorithms have been used to evolve repertoires of robot behaviours in a number of different domains, for instance: the evolution 25 of virtual walking creatures (including body plan and control policy) [12]; repertoires of morphological designs for walking soft robots [13]; repertoires of locomotion behaviours for legged robots [14][15][16][17][18][19] and four-wheeled steering robots [20]; repertoires of robotic arm behaviours [13,19,21]; repertoires of swarm behaviours [22,23]; and repertoires of controllers for maze-navigation 30 tasks [8, 17,24]. The automatic generation of behaviour repertoires resembles self-exploration or babbling in developmental robotics [25], in which a robot experiments with a wide variety of motor commands, and learns the association between those motor commands and their consequence.…”

Section: Motivation and Backgroundmentioning

confidence: 99%