|
Digital
Library of the European Council for Modelling
and Simulation |
Title: |
Automated
Tuning Of A Cellular Automata Using Parallel Asynchronous Particle Swarm
Optimisation |
Authors: |
Christoph Tholen, Lars Nolle, Tarek
El-Mihoub, Jan Dierks, Alexandra Burger, Oliver Zielinski |
Published in: |
(2019). ECMS 2019
Proceedings Edited by: Mauro Iacono, Francesco Palmieri, Marco Gribaudo,
Massimo Ficco, European Council for Modeling and Simulation. DOI: http://doi.org/10.7148/2019 ISSN:
2522-2422 (ONLINE) ISSN:
2522-2414 (PRINT) ISSN:
2522-2430 (CD-ROM) 33rd International ECMS Conference on
Modelling and Simulation,
Caserta, Italy, June 11th – June 14th, 2019 |
Citation
format: |
Christoph Tholen, Lars Nolle, Tarek El-Mihoub, Jan Dierks, Alexandra Burger, Oliver Zielinski (2019). Automated
Tuning Of A Cellular Automata Using Parallel Asynchronous Particle Swarm
Optimisation, ECMS 2019 Proceedings Edited
by: Mauro Iacono, Francesco Palmieri, Marco Gribaudo, Massimo Ficco European
Council for Modeling and Simulation. doi: 10.7148/2019-0030 |
DOI: |
https://doi.org/10.7148/2019-0030 |
Abstract: |
The long-term goal of this research is the development of a distributed autonomous low-cost platform for marine exploration. One application of such a platform could be the search for Submarine Groundwater Discharges (SGD) in a coastal environment. In order to design and to test new search strategies for such a platform, a simulation that effectively models the diffusion of groundwater discharge in shallow coastal waters is required. The simulation allows the evaluation of new search strategies without running the risk of losing expensive hardware during the field testing. In this paper a simulation based on cellular automata was adapted in order to resemble the behaviour of an existing physical model of a SGD. To speed up the optimization process, a novel adaptation of the Parallel Asynchronous Particle Swarm Optimisation (PAPSO) algorithm was proposed. Experiments showed that the novel PAPSO was able to reduce the time needed for optimisation by 69.1 %. Furthermore, the results found by PAPSO are 2.1 % better than the results of the Parallel Synchronous Particle Swarm Optimisation (PSPSO) algorithm. |
Full
text: |