摘要:針對神經(jīng)網(wǎng)絡(luò)、支持向量機(jī)等方法對數(shù)據(jù)樣本容量要求較高的問題,以及一般時間序列預(yù)測模型對最大負(fù)荷等隨機(jī)因素?cái)M合不足的問題,應(yīng)用時間序列的季節(jié)乘法模型對地區(qū)月度最大負(fù)荷做預(yù)測,并用GARCH模型對預(yù)測誤差進(jìn)行修正.用某電網(wǎng)的真實(shí)數(shù)據(jù)作案例,結(jié)果表明,誤差率僅為2%,預(yù)測精度良好.相比修正前的模型,誤差率下降0.5%,證明誤差修正模型有效.
關(guān)鍵詞:月最大負(fù)荷預(yù)測;時間序列乘法模型;GARCH模型;誤差修正
中圖分類號:TM715,F(xiàn)224 文獻(xiàn)標(biāo)識碼:A
The Multiplicative Model in Time Series and GARCH
Error Amending Model and Its Application
YANG Shang-dong1, LIU Jin-peng2, GUO Hao-chi2
(1. Research Department of Management Consulting,State Grid Energy Research Institute,Beijing 100052,China;
2. School of Economics and Management, North China Electric Power Univ, Beijing 102206, China)
Abstract: ANN and SVM forecasting models need large sample data, and the traditional time series forecasting model cannot fit sufficiently the biggest load due to random factors. And in order to overcome the shortcomings as mentioned, this paper applied the season-multiplicative model in time series to forecast the monthly peak load of region, and adopted the GARCH model to modify the forecasting error. The application results of the proposed model in a regional power grid show that the forecasting is precise, because the error rate is only 2%. And compared with the unmodified model, the new model’s error rate decreased by 0.5%.
Key words: monthly peak load forecasting; multiplicative model in time series; GARCH model; error amending
由于中長期最大負(fù)荷預(yù)測本身存在數(shù)據(jù)量比較少的特點(diǎn)[1], 因而需要大樣本的神經(jīng)網(wǎng)絡(luò)法和支持向量機(jī)等智能方法并不適用[2].相反,傳統(tǒng)的時間序列模型可較好地描述最大負(fù)荷這一隨機(jī)過程[3].但單用時間序列建模預(yù)測,因未考慮到的一些因素, 預(yù)測的殘差可能存在自回歸現(xiàn)象,故預(yù)測效果往往不理想[4].GARCH模型為自回歸條件異方差模型[5],能很好地消除預(yù)測殘差存在的自回歸現(xiàn)象[6].基于最大負(fù)荷數(shù)據(jù)的單一性、有限性以及季節(jié)性,本文將先用時間序列模型對最大負(fù)荷進(jìn)行擬合,在此基礎(chǔ)上再用GARCH模型對擬合誤差做修正,以提高預(yù)測精度.
4 結(jié) 論
1)通過實(shí)例驗(yàn)證,將時間序列乘法模型應(yīng)用在月最大負(fù)荷預(yù)測上,具有良好的擬合和預(yù)測能力.
2)用GARCH模型修正預(yù)測誤差,在原先基礎(chǔ)上消除了預(yù)測誤差的自回歸,具有良好的擬合以及預(yù)測能力.
參考文獻(xiàn)
[1] 康重慶,夏清,張伯明.電力系統(tǒng)負(fù)荷預(yù)測研究綜述與發(fā)展方向的探討[J].電力系統(tǒng)自動化,2004,28(7):1-11.
KANG Chong-qing, XIA Qing,ZHANG Bo-ming.Review of power system load forecasting and development[J].Automation of Electric Power Systems,2004,28(7):1-11.(In Chinese)
[2] 牛東曉,谷志紅,邢棉,等.基于數(shù)據(jù)挖掘的 SVM 短期負(fù)荷預(yù)測方法研究[J].中國電機(jī)工程學(xué)報,2006,26(18):6-12.
NIU Dong-xiao,GU Zhi-hong,XING Mian,et al. Study on forecasting approach to short-term load of SVM based on data mining [J].Proceedings of CSEE,2006,26(18):6-12. (In Chinese)
[3] ZHANG Xun,LAI K K, WANG Shou-yang.A new approach for crude oil price analysis based on Empirical Mode Decomposition[J].Energy Economics,2008,30(3):905-918.
[4] 徐聰穎,廖峰,陳震海.灰色組合模型在中長期電力負(fù)荷預(yù)測中的應(yīng)用[J]. 電力需求側(cè)管理,2011,13(2):20-23.
XU Cong-ying,LIAO Feng,CHEN Zhen-hai. Combination gray model in mid-term and long-term load forecasting [J]. Power DSM,2011,13(2):20-23. (In Chinese)
[5] 李媛媛,牛東曉,乞建勛. 基于因散經(jīng)驗(yàn)?zāi)J椒纸獾碾娏ω?fù)荷混合預(yù)測方法[J]. 電網(wǎng)技術(shù), 2008,32(8):58-62.
LI Yuan-yuan,NIU Dong-xiao,QI Jian-xun. A novel hybrid power load forecasting method based on ensemble empirical mode decomposition[J]. Power System Technology, 2008,32(8): 58-62. (In Chinese)
[6] 高鳳,張德生,郭熊娃. 基于相關(guān)系數(shù)的變權(quán)組合預(yù)測模型及其應(yīng)用[J]. 陜西科技大學(xué)學(xué)報, 2013,31(3):167-168.
GAO Feng, ZHANG De-sheng, GUO Xiong-wa. Variable weights combination forecast model based on the correlation coefficient and its application [J]. Journal of Shanxi University of Science and Technology 2013,31(3):167-168. (In Chinese)
[7] WU Z,HUANG N E.Ensemble empirical mode decomposition:a noise-assisted data analysis method[J].Advances in Adaptive Data Analysis, 2009, 1(1):1-41.