Detail publikace

Optimisation of Water Management Systems Using a GPU-Accelerated Differential Evolution

JAROŠ, J. MAREK, J. MENŠÍK, P.

Originální název

Optimisation of Water Management Systems Using a GPU-Accelerated Differential Evolution

Anglický název

Optimisation of Water Management Systems Using a GPU-Accelerated Differential Evolution

Jazyk

en

Originální abstrakt

The aim of this paper is to present a tool that could optimise water levels in dams to provide the inhabitants and industry with enough fresh water while keeping a reasonable flood protection and river navigability. Since the water systems may be quite complicated and the outflow plans have to be precise, we developed a GPU-accelerated algorithm based on the differential evolution to solve this task. The experimental results show that a high quality annual outflow plans can be obtained within a minute, almost 18 times faster than on a hex-core CPU.

Anglický abstrakt

The aim of this paper is to present a tool that could optimise water levels in dams to provide the inhabitants and industry with enough fresh water while keeping a reasonable flood protection and river navigability. Since the water systems may be quite complicated and the outflow plans have to be precise, we developed a GPU-accelerated algorithm based on the differential evolution to solve this task. The experimental results show that a high quality annual outflow plans can be obtained within a minute, almost 18 times faster than on a hex-core CPU.

Dokumenty

BibTex


@inproceedings{BUT119882,
  author="Jiří {Jaroš} and Jan {Marek} and Pavel {Menšík}",
  title="Optimisation of Water Management Systems Using a GPU-Accelerated Differential Evolution",
  annote="The aim of this paper is to present a tool that could optimise water levels in
dams to provide the inhabitants and industry with enough fresh water while
keeping a reasonable flood protection and river navigability. Since the water
systems may be quite complicated and the outflow plans have to be precise, we
developed a GPU-accelerated algorithm based on the differential evolution to
solve this task. The experimental results show that a high quality annual outflow
plans can be obtained within a minute, almost 18 times faster than on a hex-core
CPU.",
  address="IEEE Computer Society",
  booktitle="Proceedings of 2015 IEEE Symposium Series on Computational Intelligence",
  chapter="119882",
  doi="10.1109/SSCI.2015.266",
  edition="NEUVEDEN",
  howpublished="electronic, physical medium",
  institution="IEEE Computer Society",
  year="2015",
  month="december",
  pages="1727--1734",
  publisher="IEEE Computer Society",
  type="conference paper"
}