Reference : A competitive Predator–Prey approach to enhance surveillance by UAV swarms
Scientific journals : Article
Engineering, computing & technology : Computer science
http://hdl.handle.net/10993/49053
A competitive Predator–Prey approach to enhance surveillance by UAV swarms
English
Stolfi, Daniel H. [> >]
Brust, Matthias R. mailto [University of Luxembourg > Interdisciplinary Centre for Security, Reliability and Trust (SNT) > PCOG]
Danoy, Grégoire mailto [University of Luxembourg > Faculty of Science, Technology and Medicine (FSTM) > Department of Computer Science (DCS)]
Bouvry, Pascal mailto [University of Luxembourg > Faculty of Science, Technology and Medicine (FSTM) > Department of Computer Science (DCS)]
2021
Applied Soft Computing
111
107701
Yes (verified by ORBilu)
International
1568-4946
[en] Swarm robotics ; Computer simulation ; Mobility model ; Unmanned aerial vehicle ; Competitive coevolutionary genetic algorithm
[en] In this paper we present the competitive optimisation of a swarm of Unmanned Aerial Vehicles (UAV) protecting a restricted area from a number of intruders following a Predator–Prey approach. We propose a Competitive Coevolutionary Genetic Algorithm (CompCGA) which optimises the parameters of the UAVs (i.e. predators) to maximise the detection of intruders, while the parameters of the intruders (i.e. preys) are optimised to maximise their intrusion success rate. Having chosen the CACOC (Chaotic Ant Colony Optimisation for Coverage) as the base mobility model for the UAVs, we propose an improved new version, where its behaviour is modified by identifying and optimising new parameters to improve the overall success rate when detecting intruders. Six case studies have been optimised using simulations by performing 30 independent runs (180 in total) of our CompCGA. Finally, we conducted a series of master tournaments (1,800,000 evaluations) using the best specimens obtained from each run and case study to test the robustness of our proposed approach against unexpected intruders. Our surveillance system improved the average percentage of intruders detected with respect to CACOC by a maximum of 126%. More than 90% of intruders were detected on average when using a swarm of 16 UAVs while CACOC’s detection rates are always under 80% in all cases.
ONRG
HUNTED
http://hdl.handle.net/10993/49053
10.1016/j.asoc.2021.107701
https://www.sciencedirect.com/science/article/pii/S1568494621006220

File(s) associated to this reference

Fulltext file(s):

FileCommentaryVersionSizeAccess
Limited access
1-s2.0-S1568494621006220-main.pdfPublisher postprint3.67 MBRequest a copy

Bookmark and Share SFX Query

All documents in ORBilu are protected by a user license.