Abstract

A shortage of large power system data sets, as well as the frequent restrictions on sharing such models, have led to newfound interest in creating synthetic data that can be easily shared among researchers. This paper considers the problem of forming a power system test case from the constituent parts of realistic power grid samples. Starting from a topology, and samples of generation, load, and branches, we assemble systems while respecting the constraints imposed by a typical Optimal Power Flow problem. Expressed in this manner, the problem involves solving for permutations of the input data. Since permutations matrices are binary, the problem is linearized, allowing for a Mixed Integer Linear Problem formulation. The problem is further decomposed using the Alternating Direction Method of Multipliers as well as an Evolutionary Algorithm to facilitate scaling to larger system sizes. A post processing step is used to add shunt elements for reactive power planning. The resulting systems demonstrate statistically similar power flow behavior to reference systems. Finally, new analysis avenues, opened by synthesizing test cases according to the proposed method, are briefly introduced by creating fictitious systems with different topology models and examining how these affect power flow behavior.

Original languageEnglish (US)
JournalIEEE Transactions on Power Systems
DOIs
StateAccepted/In press - Aug 3 2018

Fingerprint

Topology
Reactive power
Evolutionary algorithms
Planning
Processing

Keywords

  • Assignment Problem
  • Geography
  • Impedance
  • Linearized Power Flow
  • Load flow
  • Load modeling
  • MILP
  • Synthetic Test Cases
  • Testing
  • Topology

ASJC Scopus subject areas

  • Energy Engineering and Power Technology
  • Electrical and Electronic Engineering

Cite this

@article{080ee179965a42ac81d201d62145ac3c,
title = "A Mathematical Programing Solution for Automatic Generation of Synthetic Power Flow Cases",
abstract = "A shortage of large power system data sets, as well as the frequent restrictions on sharing such models, have led to newfound interest in creating synthetic data that can be easily shared among researchers. This paper considers the problem of forming a power system test case from the constituent parts of realistic power grid samples. Starting from a topology, and samples of generation, load, and branches, we assemble systems while respecting the constraints imposed by a typical Optimal Power Flow problem. Expressed in this manner, the problem involves solving for permutations of the input data. Since permutations matrices are binary, the problem is linearized, allowing for a Mixed Integer Linear Problem formulation. The problem is further decomposed using the Alternating Direction Method of Multipliers as well as an Evolutionary Algorithm to facilitate scaling to larger system sizes. A post processing step is used to add shunt elements for reactive power planning. The resulting systems demonstrate statistically similar power flow behavior to reference systems. Finally, new analysis avenues, opened by synthesizing test cases according to the proposed method, are briefly introduced by creating fictitious systems with different topology models and examining how these affect power flow behavior.",
keywords = "Assignment Problem, Geography, Impedance, Linearized Power Flow, Load flow, Load modeling, MILP, Synthetic Test Cases, Testing, Topology",
author = "Eran Schweitzer and Anna Scaglione",
year = "2018",
month = "8",
day = "3",
doi = "10.1109/TPWRS.2018.2863266",
language = "English (US)",
journal = "IEEE Transactions on Power Systems",
issn = "0885-8950",
publisher = "Institute of Electrical and Electronics Engineers Inc.",

}

TY - JOUR

T1 - A Mathematical Programing Solution for Automatic Generation of Synthetic Power Flow Cases

AU - Schweitzer, Eran

AU - Scaglione, Anna

PY - 2018/8/3

Y1 - 2018/8/3

N2 - A shortage of large power system data sets, as well as the frequent restrictions on sharing such models, have led to newfound interest in creating synthetic data that can be easily shared among researchers. This paper considers the problem of forming a power system test case from the constituent parts of realistic power grid samples. Starting from a topology, and samples of generation, load, and branches, we assemble systems while respecting the constraints imposed by a typical Optimal Power Flow problem. Expressed in this manner, the problem involves solving for permutations of the input data. Since permutations matrices are binary, the problem is linearized, allowing for a Mixed Integer Linear Problem formulation. The problem is further decomposed using the Alternating Direction Method of Multipliers as well as an Evolutionary Algorithm to facilitate scaling to larger system sizes. A post processing step is used to add shunt elements for reactive power planning. The resulting systems demonstrate statistically similar power flow behavior to reference systems. Finally, new analysis avenues, opened by synthesizing test cases according to the proposed method, are briefly introduced by creating fictitious systems with different topology models and examining how these affect power flow behavior.

AB - A shortage of large power system data sets, as well as the frequent restrictions on sharing such models, have led to newfound interest in creating synthetic data that can be easily shared among researchers. This paper considers the problem of forming a power system test case from the constituent parts of realistic power grid samples. Starting from a topology, and samples of generation, load, and branches, we assemble systems while respecting the constraints imposed by a typical Optimal Power Flow problem. Expressed in this manner, the problem involves solving for permutations of the input data. Since permutations matrices are binary, the problem is linearized, allowing for a Mixed Integer Linear Problem formulation. The problem is further decomposed using the Alternating Direction Method of Multipliers as well as an Evolutionary Algorithm to facilitate scaling to larger system sizes. A post processing step is used to add shunt elements for reactive power planning. The resulting systems demonstrate statistically similar power flow behavior to reference systems. Finally, new analysis avenues, opened by synthesizing test cases according to the proposed method, are briefly introduced by creating fictitious systems with different topology models and examining how these affect power flow behavior.

KW - Assignment Problem

KW - Geography

KW - Impedance

KW - Linearized Power Flow

KW - Load flow

KW - Load modeling

KW - MILP

KW - Synthetic Test Cases

KW - Testing

KW - Topology

UR - http://www.scopus.com/inward/record.url?scp=85051019238&partnerID=8YFLogxK

UR - http://www.scopus.com/inward/citedby.url?scp=85051019238&partnerID=8YFLogxK

U2 - 10.1109/TPWRS.2018.2863266

DO - 10.1109/TPWRS.2018.2863266

M3 - Article

JO - IEEE Transactions on Power Systems

JF - IEEE Transactions on Power Systems

SN - 0885-8950

ER -