Optimal Power Dispatch Under Load Uncertainty Using a Stochastic Approximation Method

Tianqi Hong, Ashhar Raza, Francisco De Leon

Research output: Contribution to journalArticle

Abstract

This paper applies a known stochastic approximation method to solve a two-phase optimal power dispatch problem under load uncertainty using a novel realistic model for random loads. The stochastic power dispatch problem is solved with the Robbins–Monro method applied with the Kiefer–Wolfowitz procedure with random directions. The constraints of this optimization problem have been investigated and considered by truncated algorithms. Two numerical examples are presented for illustration of the significant improvement obtained with the proposed method. In the first example, the optimizable cost of the system can be reduced by 1.6% under 2.68% load variation. In the second example, the results show that the possibility of voltage violations is reduced from 49.5% to 0.01%. All the improvements are compared with deterministic optimal solutions validated with scenario-based Monte Carlo simulations. The method is 60 times faster than scenario-based Monte Carlo for similar accuracy for the IEEE 30-bus test system.

Original languageEnglish (US)
JournalIEEE Transactions on Power Systems
DOIs
StateAccepted/In press - Jan 27 2016

Fingerprint

Electric potential
Costs
Uncertainty
Monte Carlo simulation

ASJC Scopus subject areas

  • Electrical and Electronic Engineering
  • Energy Engineering and Power Technology

Cite this

Optimal Power Dispatch Under Load Uncertainty Using a Stochastic Approximation Method. / Hong, Tianqi; Raza, Ashhar; De Leon, Francisco.

In: IEEE Transactions on Power Systems, 27.01.2016.

Research output: Contribution to journalArticle

@article{7a02938410a9405cbe52b7191dd8a07d,
title = "Optimal Power Dispatch Under Load Uncertainty Using a Stochastic Approximation Method",
abstract = "This paper applies a known stochastic approximation method to solve a two-phase optimal power dispatch problem under load uncertainty using a novel realistic model for random loads. The stochastic power dispatch problem is solved with the Robbins–Monro method applied with the Kiefer–Wolfowitz procedure with random directions. The constraints of this optimization problem have been investigated and considered by truncated algorithms. Two numerical examples are presented for illustration of the significant improvement obtained with the proposed method. In the first example, the optimizable cost of the system can be reduced by 1.6{\%} under 2.68{\%} load variation. In the second example, the results show that the possibility of voltage violations is reduced from 49.5{\%} to 0.01{\%}. All the improvements are compared with deterministic optimal solutions validated with scenario-based Monte Carlo simulations. The method is 60 times faster than scenario-based Monte Carlo for similar accuracy for the IEEE 30-bus test system.",
author = "Tianqi Hong and Ashhar Raza and {De Leon}, Francisco",
year = "2016",
month = "1",
day = "27",
doi = "10.1109/TPWRS.2016.2518205",
language = "English (US)",
journal = "IEEE Transactions on Power Systems",
issn = "0885-8950",
publisher = "Institute of Electrical and Electronics Engineers Inc.",

}

TY - JOUR

T1 - Optimal Power Dispatch Under Load Uncertainty Using a Stochastic Approximation Method

AU - Hong, Tianqi

AU - Raza, Ashhar

AU - De Leon, Francisco

PY - 2016/1/27

Y1 - 2016/1/27

N2 - This paper applies a known stochastic approximation method to solve a two-phase optimal power dispatch problem under load uncertainty using a novel realistic model for random loads. The stochastic power dispatch problem is solved with the Robbins–Monro method applied with the Kiefer–Wolfowitz procedure with random directions. The constraints of this optimization problem have been investigated and considered by truncated algorithms. Two numerical examples are presented for illustration of the significant improvement obtained with the proposed method. In the first example, the optimizable cost of the system can be reduced by 1.6% under 2.68% load variation. In the second example, the results show that the possibility of voltage violations is reduced from 49.5% to 0.01%. All the improvements are compared with deterministic optimal solutions validated with scenario-based Monte Carlo simulations. The method is 60 times faster than scenario-based Monte Carlo for similar accuracy for the IEEE 30-bus test system.

AB - This paper applies a known stochastic approximation method to solve a two-phase optimal power dispatch problem under load uncertainty using a novel realistic model for random loads. The stochastic power dispatch problem is solved with the Robbins–Monro method applied with the Kiefer–Wolfowitz procedure with random directions. The constraints of this optimization problem have been investigated and considered by truncated algorithms. Two numerical examples are presented for illustration of the significant improvement obtained with the proposed method. In the first example, the optimizable cost of the system can be reduced by 1.6% under 2.68% load variation. In the second example, the results show that the possibility of voltage violations is reduced from 49.5% to 0.01%. All the improvements are compared with deterministic optimal solutions validated with scenario-based Monte Carlo simulations. The method is 60 times faster than scenario-based Monte Carlo for similar accuracy for the IEEE 30-bus test system.

UR - http://www.scopus.com/inward/record.url?scp=84961387325&partnerID=8YFLogxK

UR - http://www.scopus.com/inward/citedby.url?scp=84961387325&partnerID=8YFLogxK

U2 - 10.1109/TPWRS.2016.2518205

DO - 10.1109/TPWRS.2016.2518205

M3 - Article

AN - SCOPUS:84961387325

JO - IEEE Transactions on Power Systems

JF - IEEE Transactions on Power Systems

SN - 0885-8950

ER -