基于多代理的电力市场仿真研究
|
摘要
电力市场仿真是研究电力市场运行规律的一个非常有效的方法。近年来,基于代理的计算经济学方法在电力市场仿真研究中得到了国内外研究人员广泛的关注,并取得了较多的研究成果。然而,基于代理的市场仿真方法还有很多问题没有解决。其中,最基本的一个问题是研究方法的问题,例如,模型参数应该如何校正、仿真实验应该如何设计、仿真结果应该如何分析等。由于基于代理的仿真模型本身具有异质性和不稳定性特点再加上没有统一的研究方法,导致不同研究结果没有可比性,大大限制了基于代理的计算经济学方法在电力市场仿真研究中的发展和应用。
本文针对实验的设计方法、结果的分析方法等电力市场仿真中的问题,引入实验设计与数据处理分析理论,运用正交实验设计、对比析因实验设计、方差分析的方法,并使用SPSS(Statistical Product and Service Solutions,统计产品与服务解决方案)软件进行辅助分析。本文的主要研究工作包括:
(1)探讨了电力市场仿真的基本问题,电力市场仿真研究的主要应用方向,市场规则研究、市场环境研究、报价策略研究等方面中能够影响电力市场运行的主要因素。分析了不同的电力市场仿真研究方法。
(2)将实验设计与数据处理分析理论应用到电力市场仿真实验中,综合运用不同的实验设计方法和统计检验方法,为不同目的的电力仿真实验制定实验计划,设计实验流程。
(3)对仿真实验的结果进行了系统分析,找出了复杂的影响因素中的主要因素和次要因素,并对各影响因素的设置搭配做出了优化。
实验结果表明本文提出的研究方法能够解决电力市场仿真实验中的实验设计和数据分析的部分问题,对建立科学、系统的电力市场仿真研究实验方法有积极意义。
Power market simulation turns to be a very effective method to study on power market operating rules.In recent years, power market simulation by ACE is widely studied by foreign and domestic researchers who have done many researches on it. However, there are still many issues which have not solved, among which, analyze method is a basic problem, for example, how to correct the parameters, how to design the simulation, how to analyze the simulation results etc. the heterogeneity and instability features of the model plus no unified research methods cause different results incomparable, which greatly limits the development and application of ACE method in power market simulation.
This paper introduces experimental design and data processing analysis theory, uses the orthogonal experiment design method, variance analysis method and contrast factorial experiment design method , besides, SPSS(statistical product and service solutions) is also used as an assistant software to solve the experimental design and result analysis problems. Main research work in this paper includes:
(1) The first chapter discusses the basic problems and main application direction of power market simulation , and main factors which affect the market operation such as market rules, market environment and bidding strategy. Different simulation methods are also analyzed in this chapter.
(2) Systematized analysis is done on the simulation results to find out the main factors and secondary factors of all the complex influence factors ,and optimization is made on the setting collocation of each affecting factors.
(3) experimental design method and data processing analysis theory are applied to the simulation to draw up experiment plan and design formulated procedures by using different experiment design methods and statistic test methods.
The simulation results shows that the research method proposed in the paper can solve some problems on experiment design and data analysis, and it has positive significance in establishing scientific and systematized research method of power market simulation.
本文针对实验的设计方法、结果的分析方法等电力市场仿真中的问题,引入实验设计与数据处理分析理论,运用正交实验设计、对比析因实验设计、方差分析的方法,并使用SPSS(Statistical Product and Service Solutions,统计产品与服务解决方案)软件进行辅助分析。本文的主要研究工作包括:
(1)探讨了电力市场仿真的基本问题,电力市场仿真研究的主要应用方向,市场规则研究、市场环境研究、报价策略研究等方面中能够影响电力市场运行的主要因素。分析了不同的电力市场仿真研究方法。
(2)将实验设计与数据处理分析理论应用到电力市场仿真实验中,综合运用不同的实验设计方法和统计检验方法,为不同目的的电力仿真实验制定实验计划,设计实验流程。
(3)对仿真实验的结果进行了系统分析,找出了复杂的影响因素中的主要因素和次要因素,并对各影响因素的设置搭配做出了优化。
实验结果表明本文提出的研究方法能够解决电力市场仿真实验中的实验设计和数据分析的部分问题,对建立科学、系统的电力市场仿真研究实验方法有积极意义。
Power market simulation turns to be a very effective method to study on power market operating rules.In recent years, power market simulation by ACE is widely studied by foreign and domestic researchers who have done many researches on it. However, there are still many issues which have not solved, among which, analyze method is a basic problem, for example, how to correct the parameters, how to design the simulation, how to analyze the simulation results etc. the heterogeneity and instability features of the model plus no unified research methods cause different results incomparable, which greatly limits the development and application of ACE method in power market simulation.
This paper introduces experimental design and data processing analysis theory, uses the orthogonal experiment design method, variance analysis method and contrast factorial experiment design method , besides, SPSS(statistical product and service solutions) is also used as an assistant software to solve the experimental design and result analysis problems. Main research work in this paper includes:
(1) The first chapter discusses the basic problems and main application direction of power market simulation , and main factors which affect the market operation such as market rules, market environment and bidding strategy. Different simulation methods are also analyzed in this chapter.
(2) Systematized analysis is done on the simulation results to find out the main factors and secondary factors of all the complex influence factors ,and optimization is made on the setting collocation of each affecting factors.
(3) experimental design method and data processing analysis theory are applied to the simulation to draw up experiment plan and design formulated procedures by using different experiment design methods and statistic test methods.
The simulation results shows that the research method proposed in the paper can solve some problems on experiment design and data analysis, and it has positive significance in establishing scientific and systematized research method of power market simulation.
引文
[1] Lee S T .Simulation model explores alternative wholesale power market structures[J]·IEEE CornpuLer ApplicaLions in Power,2000. 15(1):28一35.
[2]李佳道,刘亚安,管晓宏等.电力市场仿真的系统设计及实现.空军工程大学学报(自然科学版),2003,2(4):87-90.
[3]杨晓花,周渝慧,郑亮.电力市场仿真平台研究现状综述.吉林电力,2005,5(12):50-53.
[4]荆朝霞,陈皓勇,周海明,史述红.电力市场仿真研究现状[J].高科技与产业化,2009,(03):96-99.
[5]王野平,荆朝霞,陈皓勇.基于代理的电力市场仿真的实验设计方法探讨.电力系统自化.2009,33 (17):56-60.
[6] Derek W Bunn ,Fernando S Oliveira. Agent-Based Simulation—An Application to the New Electricity Trading Arrangements of England and Wales .IEEE Transactions on Evolutionary Computation, 2001, 5(5) :493-503.
[7] Anthony J Bagnall,George D Smith. A multi-agent model of the UK market in electricity generation[J] .IEEE Transactions onEvolutionary Computation, 2005, 9 (5) :522-536
[8] Sueyoshi T,Tadiparthi G R. Agent-based approach to handle business complexity in U.S.wholesale power trading[J] .IEEE Transactions on Power Systems, 2007, 22 (2) :532-544 .
[9] Tellidou A C,Bakirtzis A G. Agent-Based Analysis of Capacity Withholding and Tacit Collusion in Electricity Markets[J] .IEEE Trans on Power Systems, 2007, 22 (4) :1735-1742 .
[10] Nicolaisen J, Petrov V, Tesfatsion L. Market power and efficiency in a computational electricity market with discriminatory double-auction pricing .IEEE Transactions on Evolutionary Computation, 2001, 5 (5) :504-523 .
[11] SUEYOSHI T,TADIPARTHI G. Wholesale power price dynamics under transmissionline li mits:a use of an agent-based intelligent si mulator .IEEE Trans on Systems,Man and Cybernetic—Part C:Applications and Reviews, 2008, 38 (2) :229-241.
[12] VEIT D J,WEIDLICH A,YAO J,et al. Si mulating the dynamics in two-settlement electricity markets via an agent-based approach .International Journal of Management Science and Engineering Management, 2006, 1 (2) :83-97 .
[13]宋依群,高瞻. Agent技术在电力市场中的应用综述.电力系统自动化学报,2008,20(3) :111-115.
[14]邹斌.李庆华.言茂松.电力拍卖市场的智能代理模拟模型.国电机工程学报.2005, 25( 15) : 7- 1 1.
[15]陈皓勇,杨彦,张尧.电力市场智能模拟中代理决策模块的实现.电力系统自动化,2008, 32 (20) :12-16.
[16]康重庆,江健健,夏清.基于智能个体信念学习的电力市场模拟的理论框架[J].电网技术, 2005,(12) .
[17] TESFATSION.L.Experimental.design;basic.concepts.and.terminology[EB/OL] .http://www.econ.iastate.edu/tesfatsi/empvalid.ht m#Design. [2009-05-02].
[18]刘梅招,杨莉,甘德强.基于Agent的电力市场仿真研究综述[J].电网技术, 2005,(04):76-80.
[19]王威,王雨,王勇.基于Agent的电力市场价格仿真研究[J].微计算机信息, 2007,(06),26-28.
[20]张志良,楚丰,游大海,龙云.一种基于CORBA和Agent技术的电力市场仿真平台的研究[J].电力系统及其自动化学报, 2002,(04),29-31.
[21]宋依群,吴炯.基于Q学习算法的发电公司决策新模型.上海交通大学学报, 2006,40 (4) :568-571.
[22]王小平,曹立明.遗传算法理论、应用与软件实现[M].西安:西安交通大学出版社,2004.
[23]李运灵,侯云鹤等.遗传算法在电力市场中的应用.电网技术,2000,24(3):55-58.
[24]李运灵,侯云鹤等.基于遗传算法的电力市场投标竞价模型.电力系统及其自动化学报,2000,12(6):45-48.
[25]陈皓勇,王锡凡,别朝红等.协同进化算法及其在电力系统中的应用前景[J].电力系统自动化,2003,27(23):94-100.
[26]杨彦,陈皓勇,张尧等.基于协同进化算法求解寡头电力市场均衡[J].电力系统自动化,2009,33(18):42-46.
[27]陈皓勇,王锡凡,王秀丽等.基于Java的电力市场竞价实验平台设计、实现及应用[J].电力系统自动化,2004,28(17):22-26.
[28] Smith V L. Regulatory Reform in the Electric Power Industry .Regulation, 1996, 19(1) :33-46 .
[29] Davis D D,Holt C A.Experimental Economics.Princeton(NJ):Princeton University Press, 1993.
[30] Kagel.J.H,Roth.A.E.The.Handbook.of.ExperimentalEconomics.Princeton(NJ):Princeton University Press,1995.
[31] Rassenti S J, Smith V L,Wilson B J. Using Experiments to Inform the Privatization/Deregulation Movement in Electricity .Cato Journal, 2002, 21(3) :515-544 .
[32] Smith V L,Rassenti S J,Wilson B J. California:Energy Crisis or Market Design Crisis?http://www. ices-gmu. net/article.php/373. html
[33] Smith V L, Rassenti S J, Wilson B J. California: energy crisis or market design crisis [R]. http://www-hoover.stanford.edu/research/ conferences/calelectricity/smith.pdf, 2001.
[34] Rassenti S J, Smith V L, Wilson B J. Turning off the Lights .Regulation, 2001, 24(3) :70-76 .
[35] Rassenti S J, Smith V L, Wilson B J.Turning off the Lights.Regulation,2001,24(3):70-76.
[36] Rassenti S J, Smith V L, Wilson B J. Using experiments to inform the privatization/deregulation movement in electricity [J]. Cato Journal, 2002, 21(3): 515-544.
[37] Rassenti S J, Smith V L, Wilson B J. Discriminatory price auctions in electricity markets: low volatility at the expense of high price levels [J]. Journal of Regulatory Economics, 2003, 23(2): 109-123.
[38]刘振学,黄仁和,田爱民等.实验设计与数据处理[M]:化学工业出版社,2005.
[39]威勒(Willer, David)沃克(Walker, Henry A.),实验设计原理:社会科学理论验证的一种路径[M].杜伟宇,孟琦:重庆大学出版社,1943。
[40]方开泰,马长兴等.正交与均匀实验设计[M]:科学出版社,2001
[41]荆朝霞,杨莹,颜汉荣等.随机数对基于代理的电力市场仿真实验的影响分析.电力系统自动化.2010, 34(14):58-62, 83
[42]荆朝霞,杨莹. EWA算法在电力市场仿真中的应用[J].电力系统自动化, 2010,(24),48-50.
[43] Camerer, Colin F. , Teck H-Ho. EWA learning in Normal-form Games [J ] . Econometrica , 1999 , 67 : 827 -873.
[2]李佳道,刘亚安,管晓宏等.电力市场仿真的系统设计及实现.空军工程大学学报(自然科学版),2003,2(4):87-90.
[3]杨晓花,周渝慧,郑亮.电力市场仿真平台研究现状综述.吉林电力,2005,5(12):50-53.
[4]荆朝霞,陈皓勇,周海明,史述红.电力市场仿真研究现状[J].高科技与产业化,2009,(03):96-99.
[5]王野平,荆朝霞,陈皓勇.基于代理的电力市场仿真的实验设计方法探讨.电力系统自化.2009,33 (17):56-60.
[6] Derek W Bunn ,Fernando S Oliveira. Agent-Based Simulation—An Application to the New Electricity Trading Arrangements of England and Wales .IEEE Transactions on Evolutionary Computation, 2001, 5(5) :493-503.
[7] Anthony J Bagnall,George D Smith. A multi-agent model of the UK market in electricity generation[J] .IEEE Transactions onEvolutionary Computation, 2005, 9 (5) :522-536
[8] Sueyoshi T,Tadiparthi G R. Agent-based approach to handle business complexity in U.S.wholesale power trading[J] .IEEE Transactions on Power Systems, 2007, 22 (2) :532-544 .
[9] Tellidou A C,Bakirtzis A G. Agent-Based Analysis of Capacity Withholding and Tacit Collusion in Electricity Markets[J] .IEEE Trans on Power Systems, 2007, 22 (4) :1735-1742 .
[10] Nicolaisen J, Petrov V, Tesfatsion L. Market power and efficiency in a computational electricity market with discriminatory double-auction pricing .IEEE Transactions on Evolutionary Computation, 2001, 5 (5) :504-523 .
[11] SUEYOSHI T,TADIPARTHI G. Wholesale power price dynamics under transmissionline li mits:a use of an agent-based intelligent si mulator .IEEE Trans on Systems,Man and Cybernetic—Part C:Applications and Reviews, 2008, 38 (2) :229-241.
[12] VEIT D J,WEIDLICH A,YAO J,et al. Si mulating the dynamics in two-settlement electricity markets via an agent-based approach .International Journal of Management Science and Engineering Management, 2006, 1 (2) :83-97 .
[13]宋依群,高瞻. Agent技术在电力市场中的应用综述.电力系统自动化学报,2008,20(3) :111-115.
[14]邹斌.李庆华.言茂松.电力拍卖市场的智能代理模拟模型.国电机工程学报.2005, 25( 15) : 7- 1 1.
[15]陈皓勇,杨彦,张尧.电力市场智能模拟中代理决策模块的实现.电力系统自动化,2008, 32 (20) :12-16.
[16]康重庆,江健健,夏清.基于智能个体信念学习的电力市场模拟的理论框架[J].电网技术, 2005,(12) .
[17] TESFATSION.L.Experimental.design;basic.concepts.and.terminology[EB/OL] .http://www.econ.iastate.edu/tesfatsi/empvalid.ht m#Design. [2009-05-02].
[18]刘梅招,杨莉,甘德强.基于Agent的电力市场仿真研究综述[J].电网技术, 2005,(04):76-80.
[19]王威,王雨,王勇.基于Agent的电力市场价格仿真研究[J].微计算机信息, 2007,(06),26-28.
[20]张志良,楚丰,游大海,龙云.一种基于CORBA和Agent技术的电力市场仿真平台的研究[J].电力系统及其自动化学报, 2002,(04),29-31.
[21]宋依群,吴炯.基于Q学习算法的发电公司决策新模型.上海交通大学学报, 2006,40 (4) :568-571.
[22]王小平,曹立明.遗传算法理论、应用与软件实现[M].西安:西安交通大学出版社,2004.
[23]李运灵,侯云鹤等.遗传算法在电力市场中的应用.电网技术,2000,24(3):55-58.
[24]李运灵,侯云鹤等.基于遗传算法的电力市场投标竞价模型.电力系统及其自动化学报,2000,12(6):45-48.
[25]陈皓勇,王锡凡,别朝红等.协同进化算法及其在电力系统中的应用前景[J].电力系统自动化,2003,27(23):94-100.
[26]杨彦,陈皓勇,张尧等.基于协同进化算法求解寡头电力市场均衡[J].电力系统自动化,2009,33(18):42-46.
[27]陈皓勇,王锡凡,王秀丽等.基于Java的电力市场竞价实验平台设计、实现及应用[J].电力系统自动化,2004,28(17):22-26.
[28] Smith V L. Regulatory Reform in the Electric Power Industry .Regulation, 1996, 19(1) :33-46 .
[29] Davis D D,Holt C A.Experimental Economics.Princeton(NJ):Princeton University Press, 1993.
[30] Kagel.J.H,Roth.A.E.The.Handbook.of.ExperimentalEconomics.Princeton(NJ):Princeton University Press,1995.
[31] Rassenti S J, Smith V L,Wilson B J. Using Experiments to Inform the Privatization/Deregulation Movement in Electricity .Cato Journal, 2002, 21(3) :515-544 .
[32] Smith V L,Rassenti S J,Wilson B J. California:Energy Crisis or Market Design Crisis?http://www. ices-gmu. net/article.php/373. html
[33] Smith V L, Rassenti S J, Wilson B J. California: energy crisis or market design crisis [R]. http://www-hoover.stanford.edu/research/ conferences/calelectricity/smith.pdf, 2001.
[34] Rassenti S J, Smith V L, Wilson B J. Turning off the Lights .Regulation, 2001, 24(3) :70-76 .
[35] Rassenti S J, Smith V L, Wilson B J.Turning off the Lights.Regulation,2001,24(3):70-76.
[36] Rassenti S J, Smith V L, Wilson B J. Using experiments to inform the privatization/deregulation movement in electricity [J]. Cato Journal, 2002, 21(3): 515-544.
[37] Rassenti S J, Smith V L, Wilson B J. Discriminatory price auctions in electricity markets: low volatility at the expense of high price levels [J]. Journal of Regulatory Economics, 2003, 23(2): 109-123.
[38]刘振学,黄仁和,田爱民等.实验设计与数据处理[M]:化学工业出版社,2005.
[39]威勒(Willer, David)沃克(Walker, Henry A.),实验设计原理:社会科学理论验证的一种路径[M].杜伟宇,孟琦:重庆大学出版社,1943。
[40]方开泰,马长兴等.正交与均匀实验设计[M]:科学出版社,2001
[41]荆朝霞,杨莹,颜汉荣等.随机数对基于代理的电力市场仿真实验的影响分析.电力系统自动化.2010, 34(14):58-62, 83
[42]荆朝霞,杨莹. EWA算法在电力市场仿真中的应用[J].电力系统自动化, 2010,(24),48-50.
[43] Camerer, Colin F. , Teck H-Ho. EWA learning in Normal-form Games [J ] . Econometrica , 1999 , 67 : 827 -873.