Permutation-based optimization for the load restoration problem with improved time estimation of maneuvers

Fillipe Goulart, André L. Maravilha, Eduardo G. Carrano, Felipe Campelo

Research output: Contribution to journalArticle

Abstract

After the occurrence of faults in a radial distribution system, the load restoration problem consists in implementing a sequence of switch opening and closing operations such that the resulting network configuration restores services to the most loads in the shortest possible time. We formulate this optimization problem in terms of two complementary objectives, minimizing simultaneously the energy not supplied and the power not restored. The search space is encoded as a set of permutation vectors containing all maneuverable switches, and the decoding mechanism always guarantees feasibility and allows for multiple solutions per vector. In order to cope with the possibly large search space, an efficient reduction mechanism is proposed to decrease the number of allowed permutations. The resulting optimization problem is solved using Simulated Annealing followed by a local search refinement. The time taken to perform the maneuvers is estimated using a scheduling approach, which takes into account the existence of multiple dispatch teams and thus provides a more reliable computation than the usual approach of using the number of switch operations. The proposed method is validated using known optimal results in small problem instances, and is able to return significantly better results when compared against a Branch and Bound method with a pruning heuristic in a more complex scenario.
Original languageEnglish
Pages (from-to)339-355
Number of pages16
JournalInternational Journal of Electrical Power & Energy Systems
Volume101
Early online date6 Apr 2018
DOIs
Publication statusPublished - 1 Oct 2018

Fingerprint

Restoration
Switches
Branch and bound method
Plant shutdowns
Simulated annealing
Decoding
Scheduling
Local search (optimization)

Keywords

  • Load restoration
  • meta-heuristics
  • multi-objective optimization

Cite this

Goulart, F., Maravilha, A. L., Carrano, E. G., & Campelo, F. (2018). Permutation-based optimization for the load restoration problem with improved time estimation of maneuvers. International Journal of Electrical Power & Energy Systems, 101, 339-355. https://doi.org/10.1016/j.ijepes.2018.03.030
Goulart, Fillipe ; Maravilha, André L. ; Carrano, Eduardo G. ; Campelo, Felipe. / Permutation-based optimization for the load restoration problem with improved time estimation of maneuvers. In: International Journal of Electrical Power & Energy Systems. 2018 ; Vol. 101. pp. 339-355.
@article{3b47cc348c98424ea1ff02338bd48fbc,
title = "Permutation-based optimization for the load restoration problem with improved time estimation of maneuvers",
abstract = "After the occurrence of faults in a radial distribution system, the load restoration problem consists in implementing a sequence of switch opening and closing operations such that the resulting network configuration restores services to the most loads in the shortest possible time. We formulate this optimization problem in terms of two complementary objectives, minimizing simultaneously the energy not supplied and the power not restored. The search space is encoded as a set of permutation vectors containing all maneuverable switches, and the decoding mechanism always guarantees feasibility and allows for multiple solutions per vector. In order to cope with the possibly large search space, an efficient reduction mechanism is proposed to decrease the number of allowed permutations. The resulting optimization problem is solved using Simulated Annealing followed by a local search refinement. The time taken to perform the maneuvers is estimated using a scheduling approach, which takes into account the existence of multiple dispatch teams and thus provides a more reliable computation than the usual approach of using the number of switch operations. The proposed method is validated using known optimal results in small problem instances, and is able to return significantly better results when compared against a Branch and Bound method with a pruning heuristic in a more complex scenario.",
keywords = "Load restoration, meta-heuristics, multi-objective optimization",
author = "Fillipe Goulart and Maravilha, {Andr{\'e} L.} and Carrano, {Eduardo G.} and Felipe Campelo",
year = "2018",
month = "10",
day = "1",
doi = "10.1016/j.ijepes.2018.03.030",
language = "English",
volume = "101",
pages = "339--355",

}

Goulart, F, Maravilha, AL, Carrano, EG & Campelo, F 2018, 'Permutation-based optimization for the load restoration problem with improved time estimation of maneuvers', International Journal of Electrical Power & Energy Systems, vol. 101, pp. 339-355. https://doi.org/10.1016/j.ijepes.2018.03.030

Permutation-based optimization for the load restoration problem with improved time estimation of maneuvers. / Goulart, Fillipe; Maravilha, André L.; Carrano, Eduardo G.; Campelo, Felipe.

In: International Journal of Electrical Power & Energy Systems, Vol. 101, 01.10.2018, p. 339-355.

Research output: Contribution to journalArticle

TY - JOUR

T1 - Permutation-based optimization for the load restoration problem with improved time estimation of maneuvers

AU - Goulart, Fillipe

AU - Maravilha, André L.

AU - Carrano, Eduardo G.

AU - Campelo, Felipe

PY - 2018/10/1

Y1 - 2018/10/1

N2 - After the occurrence of faults in a radial distribution system, the load restoration problem consists in implementing a sequence of switch opening and closing operations such that the resulting network configuration restores services to the most loads in the shortest possible time. We formulate this optimization problem in terms of two complementary objectives, minimizing simultaneously the energy not supplied and the power not restored. The search space is encoded as a set of permutation vectors containing all maneuverable switches, and the decoding mechanism always guarantees feasibility and allows for multiple solutions per vector. In order to cope with the possibly large search space, an efficient reduction mechanism is proposed to decrease the number of allowed permutations. The resulting optimization problem is solved using Simulated Annealing followed by a local search refinement. The time taken to perform the maneuvers is estimated using a scheduling approach, which takes into account the existence of multiple dispatch teams and thus provides a more reliable computation than the usual approach of using the number of switch operations. The proposed method is validated using known optimal results in small problem instances, and is able to return significantly better results when compared against a Branch and Bound method with a pruning heuristic in a more complex scenario.

AB - After the occurrence of faults in a radial distribution system, the load restoration problem consists in implementing a sequence of switch opening and closing operations such that the resulting network configuration restores services to the most loads in the shortest possible time. We formulate this optimization problem in terms of two complementary objectives, minimizing simultaneously the energy not supplied and the power not restored. The search space is encoded as a set of permutation vectors containing all maneuverable switches, and the decoding mechanism always guarantees feasibility and allows for multiple solutions per vector. In order to cope with the possibly large search space, an efficient reduction mechanism is proposed to decrease the number of allowed permutations. The resulting optimization problem is solved using Simulated Annealing followed by a local search refinement. The time taken to perform the maneuvers is estimated using a scheduling approach, which takes into account the existence of multiple dispatch teams and thus provides a more reliable computation than the usual approach of using the number of switch operations. The proposed method is validated using known optimal results in small problem instances, and is able to return significantly better results when compared against a Branch and Bound method with a pruning heuristic in a more complex scenario.

KW - Load restoration

KW - meta-heuristics

KW - multi-objective optimization

UR - https://www.sciencedirect.com/science/article/pii/S0142061517327588

U2 - 10.1016/j.ijepes.2018.03.030

DO - 10.1016/j.ijepes.2018.03.030

M3 - Article

VL - 101

SP - 339

EP - 355

ER -

Goulart F, Maravilha AL, Carrano EG, Campelo F. Permutation-based optimization for the load restoration problem with improved time estimation of maneuvers. International Journal of Electrical Power & Energy Systems. 2018 Oct 1;101:339-355. https://doi.org/10.1016/j.ijepes.2018.03.030