RUS  ENG
Full version
JOURNALS // Proceedings of the Institute for System Programming of the RAS // Archive

Proceedings of ISP RAS, 2021 Volume 33, Issue 2, Pages 125–136 (Mi tisp589)

Smart grid demand response strategies for datacenters

J. Murañaa, S. Nesmachnowab, S. Iturriagaa, S. Montes de Ocaa, G. Belcredia, P. Monzóna, V. Shepelevb, A. Tchernykhcdb

a Universidad de la Republica
b South Ural State University
c Centro de Investigación Científica y de Educación Superior
d Ivannikov Institute for System Programming of the Russian Academy of Sciences

Abstract: This article presents demand response techniques for the participation of datacenters in smart electricity markets under the smart grid paradigm. The proposed approach includes a datacenter model based on empirical information to determine the power consumption of CPU-intensive and memory-intensive tasks. A negotiation approach between the datacenter and clients and a heuristic planning method for energy reduction optimization are proposed. The experimental evaluation is performed over realistic problem instances modeling different types of clients. Results indicate that the proposed approach is effective to provide appropriate demand response actions according to monetary incentives.

Keywords: smart grid, computational intelligence, demand response.

DOI: 10.15514/ISPRAS-2021-33(2)-7



© Steklov Math. Inst. of RAS, 2024