A comparative study on pricing rules and its effect on total dispatch cost

Zhigang Liao, Ly Sugianto

Research output: Contribution to journalArticleResearchpeer-review

Abstract

A focus of study in auction market design is revenue comparison. Revenue Equivalence Theorem implies that bidders receive the same amount of revenue invariant to the auction type. This paper explores the applicability of this theorem in the context of the electricity market. We develop experimental test cases using agent-based simulation to examine the impact of different pricing rule on total dispatch cost. Using Q-learning in a repetitive trading environment, generator agents can quickly learn the market characteristics and seek to maximize their revenue by adapting their bidding strategies. A look-up table is utilized as a learning mechanism to improve the agent s bidding strategies. We conclude that Revenue Equivalence Theorem holds in a multi-unitmulti- period in symmetric bidding. In asymmetric bidding and when the market share of generator agents differs significantly, the computer simulation allows us to observe the rapid increase in revenue with uniform auction.
Original languageEnglish
Pages (from-to)505-510
Number of pages6
JournalProcedia Computer Science
Volume6
DOIs
Publication statusPublished - 2011

Keywords

  • Vickrey auction
  • Dispatch cost
  • Competitive market
  • Strategic bidding
  • Electricity market

Cite this

@article{952c6d8021c3439a8d0d5c70a7972677,
title = "A comparative study on pricing rules and its effect on total dispatch cost",
abstract = "A focus of study in auction market design is revenue comparison. Revenue Equivalence Theorem implies that bidders receive the same amount of revenue invariant to the auction type. This paper explores the applicability of this theorem in the context of the electricity market. We develop experimental test cases using agent-based simulation to examine the impact of different pricing rule on total dispatch cost. Using Q-learning in a repetitive trading environment, generator agents can quickly learn the market characteristics and seek to maximize their revenue by adapting their bidding strategies. A look-up table is utilized as a learning mechanism to improve the agent s bidding strategies. We conclude that Revenue Equivalence Theorem holds in a multi-unitmulti- period in symmetric bidding. In asymmetric bidding and when the market share of generator agents differs significantly, the computer simulation allows us to observe the rapid increase in revenue with uniform auction.",
keywords = "Vickrey auction, Dispatch cost, Competitive market, Strategic bidding, Electricity market",
author = "Zhigang Liao and Ly Sugianto",
year = "2011",
doi = "10.1016/j.procs.2011.08.093",
language = "English",
volume = "6",
pages = "505--510",
journal = "Procedia Computer Science",
issn = "1877-0509",
publisher = "Elsevier",

}

A comparative study on pricing rules and its effect on total dispatch cost. / Liao, Zhigang; Sugianto, Ly.

In: Procedia Computer Science, Vol. 6, 2011, p. 505-510.

Research output: Contribution to journalArticleResearchpeer-review

TY - JOUR

T1 - A comparative study on pricing rules and its effect on total dispatch cost

AU - Liao, Zhigang

AU - Sugianto, Ly

PY - 2011

Y1 - 2011

N2 - A focus of study in auction market design is revenue comparison. Revenue Equivalence Theorem implies that bidders receive the same amount of revenue invariant to the auction type. This paper explores the applicability of this theorem in the context of the electricity market. We develop experimental test cases using agent-based simulation to examine the impact of different pricing rule on total dispatch cost. Using Q-learning in a repetitive trading environment, generator agents can quickly learn the market characteristics and seek to maximize their revenue by adapting their bidding strategies. A look-up table is utilized as a learning mechanism to improve the agent s bidding strategies. We conclude that Revenue Equivalence Theorem holds in a multi-unitmulti- period in symmetric bidding. In asymmetric bidding and when the market share of generator agents differs significantly, the computer simulation allows us to observe the rapid increase in revenue with uniform auction.

AB - A focus of study in auction market design is revenue comparison. Revenue Equivalence Theorem implies that bidders receive the same amount of revenue invariant to the auction type. This paper explores the applicability of this theorem in the context of the electricity market. We develop experimental test cases using agent-based simulation to examine the impact of different pricing rule on total dispatch cost. Using Q-learning in a repetitive trading environment, generator agents can quickly learn the market characteristics and seek to maximize their revenue by adapting their bidding strategies. A look-up table is utilized as a learning mechanism to improve the agent s bidding strategies. We conclude that Revenue Equivalence Theorem holds in a multi-unitmulti- period in symmetric bidding. In asymmetric bidding and when the market share of generator agents differs significantly, the computer simulation allows us to observe the rapid increase in revenue with uniform auction.

KW - Vickrey auction

KW - Dispatch cost

KW - Competitive market

KW - Strategic bidding

KW - Electricity market

UR - http://www.scopus.com/inward/record.url?scp=84856460577&partnerID=8YFLogxK

U2 - 10.1016/j.procs.2011.08.093

DO - 10.1016/j.procs.2011.08.093

M3 - Article

VL - 6

SP - 505

EP - 510

JO - Procedia Computer Science

JF - Procedia Computer Science

SN - 1877-0509

ER -