A Forecast Method for Electricity Spot Market Clearing Prices Based on Fractal Theory

doi:10.11930/j.issn.1004-9649.202502005

Abstract

Abstract:

To address the complex fluctuations of electricity spot market clearing prices with high proportion integration of new energy, this paper proposes a clearing price prediction method based on fractal theory. Firstly, the clearing price fluctuations are analyzed from monofractal and multifractal dimensions, and corresponding prediction models are constructed. Secondly, based on the monofractal prediction method, the long-term memory and self-similarity of clearing prices are quantified using the Hurst indices and box dimension; based on multifractal method, the local fluctuation anomalies caused by sudden changes in new energy output are effectively captured via pattern matching. Thirdly, the proposed method is verified with clearing price data from five regions, which shows that, compared to the traditional "ARIMA + neural network" approach, the proposed method significantly improves the prediction accuracy, with the maximum error decreasing from 43.95% to 8.41%. In-depth applicability scenario analysis indicates that, when the Hurst index is large, the monofractal-based prediction method is more applicable; when multifractal features dominate, the multifractal-based prediction method is preferable.

Key words: new energy, spot market, clearing electricity price, fractal theory, monofractal, multifractal

WU Wenzu, WANG Xuhui, LU Yuan, CHEN Wan, TIAN Shijin, CHEN Xian. A Forecast Method for Electricity Spot Market Clearing Prices Based on Fractal Theory[J]. Electric Power, 2025, 58(9): 183-193.

Add to citation manager EndNote|Ris|BibTeX

URL: https://www.electricpower.com.cn/EN/10.11930/j.issn.1004-9649.202502005

https://www.electricpower.com.cn/EN/Y2025/V58/I9/183

Figures/Tables 9

References 25

1	国家能源局. 关于做好新能源消纳工作保障新能源高质量发展的通知[R/OL]. [2024-05-18]. http://zfxxgk.nea.gov.cn/2024-05/28/c_1310777105.htm.
2	国家能源局, 国家发展改革委. 关于加快建设全国统一电力市场体系的指导意见[R/OL]. [2022-01-28]. https://www.ndrc.gov.cn/xxgk/zcfb/tz/202201/t20220128_1313653.html.
3	莫凯佳, 何英静, 陈浔俊, 等. 基于演化博弈的清洁能源聚合发电单元竞价策略[J]. 南方电网技术, 2025, 19 (5): 83- 92.
	MO Kaijia, HE Yingjing, CHEN Xunjun, et al. Evolutionary game based bidding strategy of aggregated generation unit for renewable energy[J]. Southern Power System Technology, 2025, 19 (5): 83- 92.
4	翟哲, 陈梓煜, 刘起兴, 等. 计及风险管理的分布式资源聚合商电力市场交易模型[J]. 中国电力, 2025, 58 (6): 56- 66, 155.
	ZHAI Zhe, CHEN Ziyu, LIU Qixing, et al. An electricity market trading model for distributed resource aggregators considering risk management[J]. Electric Power, 2025, 58 (6): 56- 66, 155.
5	李咸善, 丁胜彪, 李飞, 等. 考虑水电调节费用补偿的风光水联盟优化调度策略[J]. 中国电力, 2024, 57 (5): 26- 38. DOI
	LI Xianshan, DING Shengbiao, LI Fei, et al. Optimal scheduling strategy for wind-solar-hydro alliance considering compensation of regulation by hydropower[J]. Electric Power, 2024, 57 (5): 26- 38. DOI
6	张诗若. 仿真系统中阻塞状态的电价预测[J]. 能源科技, 2024, 22 (2): 3- 7.
	ZHANG Shiruo. Electricity price forecasting for blockage status in simulation system[J]. Energy Science and Technology, 2024, 22 (2): 3- 7.
7	CONTRERAS J, ESPINOLA R, NOGALES F J, et al. ARIMA models to predict next-day electricity prices[J]. IEEE Transactions on Power Systems, 2003, 18 (3): 1014- 1020. DOI
8	徐可琪. 高比例新能源下的市场电价预测及风险评估与分析[D]. 北京: 华北电力大学, 2023.
	XU Keqi. Electricity market prices forecasting and risk assessment and analysis under high proportion of renewable energy[D]. Beijing: North China Electric Power University, 2023.
9	方逸飞, 刘东, 陈静, 等. 考虑中长期市场因素的净现货竞价空间驱动的日前出清电价预测方法[J/OL]. 中国电机工程学报, 2025: 1–13. (2025-02-25). https://kns.cnki.net/KCMS/detail/detail.aspx?filename=ZGDC20250224003&dbname=CJFD&dbcode=CJFQ.
	FANG Yifei, LIU Dong, CHEN Jing, et al. Day-ahead electricity price forecasting method driven by net spot bidding space considering medium and long-term market factors[J/OL]. Proceedings of the CSEE, 2025: 1–13. (2025-02-25). https://kns.cnki.net/KCMS/detail/detail.aspx?filename=ZGDC20250224003&dbname=CJFD&dbcode=CJFQ.
10	牛东晓, 刘达, 冯义, 等. 考虑外生变量的广义自回归条件异方差日前电价预测模型[J]. 电网技术, 2007, 31 (22): 44- 48.
	NIU Dongxiao, LIU Da, FENG Yi, et al. Exogenous variable considered generalized autoregressive conditional heteroscedastic model for day-ahead electricity price forecasting[J]. Power System Technology, 2007, 31 (22): 44- 48.
11	胡峰, 彭力. 基于时间序列模型的电价预测方法[J]. 继电器, 2008, 36 (2): 35- 40.
	HU Feng, PENG Li. Electricity price forecasting solution based on time series models[J]. Relay, 2008, 36 (2): 35- 40.
12	卢凯灵, 李盼扉, 何振锋, 等. 基于Blending集成学习模型的电力市场日前出清电价预测[J]. 电力需求侧管理, 2023, 25 (3): 27- 32. DOI
	LU Kailing, LI Panfei, HE Zhenfeng, et al. Day ahead clearing price forecasting in power market based on blending ensemble-learning model[J]. Power Demand Side Management, 2023, 25 (3): 27- 32. DOI
13	许越, 李强, 崔晖. 基于MIC-EEMD-改进Informer的含高比例清洁能源与储能的电力市场短期电价多步预测[J]. 电网技术, 2024, 48 (3): 949- 958.
	XU Yue, LI Qiang, CUI Hui. Short-term multi-step price prediction for the electricity market with a high proportion of clean energy and energy storage based on MIC-EEMD-improved informer[J]. Power System Technology, 2024, 48 (3): 949- 958.
14	郭雪丽, 华大鹏, 包鹏宇, 等. 一种基于改进VMD-PSO-CNN-LSTM的短期电价预测方法[J]. 电力科学与技术学报, 2024, 39 (2): 35- 43.
	GUO Xueli, HUA Dapeng, BAO Pengyu, et al. A short-term electricity price forecasting method based on improved VMD-PSO-CNN-LSTM[J]. Journal of Electric Power Science and Technology, 2024, 39 (2): 35- 43.
15	罗安. 基于多重分形与非均衡样本的金融市场极端风险预警研究[D]. 成都: 西南财经大学, 2023.
	LUO An. Research on early warning of extreme risks in financial markets based on multifractality and unbalanced samples[D]. Chengdu: Southwestern University of Finance and Economics, 2023.
16	贺洪增. 基于复杂网络的股票市场稳定措施和板块预测研究[D]. 北京: 北京科技大学, 2023.
	HE Hongzeng. Research on stabilization measures and prediction method of stock market based on complex network[D]. Beijing: University of Science and Technology Beijing, 2023.
17	范常容. 基于多重分形的国际油价预测研究[D]. 北京: 北京化工大学, 2019.
	FAN Changrong. Multifractal-based international oil price forecasting. Beijing: Beijing University of Chemical Technology, 2019.
18	余乐安, 范常容, 马月明. 基于混合多分形小波的国际油价多期预测研究[J]. 计量经济学报, 2021, 1 (3): 612- 623. DOI
	YU Lean, FAN Changrong, MA Yueming. Multi-period prediction of oil price with a hybrid multifractal wavelet model[J]. China Journal of Econometrics, 2021, 1 (3): 612- 623. DOI
19	杨小力. 我国碳市场有效性及波动率预测研究: 基于分形市场理论[D]. 济南: 山东师范大学, 2021.
	YANG Xiaoli. Research on the effectiveness and volatility forecast of carbon market in China: based on fractal market hypothesis. Jinan: Shandong Normal University, 2021.
20	张凯琳. 基于分形理论的我国碳排放权交易市场风险测度研究[D]. 北京: 华北电力大学, 2023.
	ZHANG Kailin. Research on risk measurement of carbon emission trading market in China based on fractal theory[D]. Beijing: North China Electric Power University, 2023.
21	孙晶琪, 牛东晓, 冷媛, 等. 电力市场复杂特性及其演化规律的实证研究[J]. 技术经济与管理研究, 2012, (11): 3- 8. DOI
	SUN Jingqi, NIU Dongxiao, LENG Yuan, et al. Empirical study of the complex characteristics of the electricity market and its evolution[J]. Technoeconomics & Management Research, 2012, (11): 3- 8. DOI
22	孙可, 韩祯祥. 美国加州电力市场电价的分形特征[J]. 能源工程, 2006, 26 (5): 1- 3. DOI
	SUN Ke, HAN Zhenxiang. Fractal feature of prices in California electricity market[J]. Energy Engineering, 2006, 26 (5): 1- 3. DOI
23	肖瑶. 基于分形理论的电力短期负荷预测研究[D]. 长沙: 中南大学, 2007.
24	张济忠. 分形[M]. 北京: 清华大学出版社, 1995: 14–18.
25	谭忠富, 张金良. 利用多因素小波变换和多变量时间序列模型的日前电价预测[J]. 中国电机工程学报, 2010, 30 (1): 103- 110.
	TAN Zhongfu, ZHANG Jinliang. Day-ahead electricity price forecasting based on multi-factor wavelet analysis and multivariate time series models[J]. Proceedings of the CSEE, 2010, 30 (1): 103- 110.

地区	$ \lambda $/%	$ h $	$ \Delta \alpha $	$ {f\left(\alpha \right)}_{\min} $	$ {f\left(\alpha \right)}_{\mathrm{m}\mathrm{a}\mathrm{x}} $	$ \Delta f\left(\alpha \right) $
A	58	0.9408	1.0116	0.6571	1.1770	0.5199
B	12	0.9846	0.4947	0.8618	1.1261	0.2643
C	32	0.8961	0.8423	0.6720	1.1263	0.4544
D	29	0.9805	0.8137	0.6802	1.0977	0.4147
E	63	0.9970	0.8708	0.6770	1.1493	0.4723

地区	$ \lambda $/%	$ h $	$ \Delta \alpha $	$ {f\left(\alpha \right)}_{\min} $	$ {f\left(\alpha \right)}_{\mathrm{m}\mathrm{a}\mathrm{x}} $	$ \Delta f\left(\alpha \right) $
A	58	0.9408	1.0116	0.6571	1.1770	0.5199
B	12	0.9846	0.4947	0.8618	1.1261	0.2643
C	32	0.8961	0.8423	0.6720	1.1263	0.4544
D	29	0.9805	0.8137	0.6802	1.0977	0.4147
E	63	0.9970	0.8708	0.6770	1.1493	0.4723

预测方法	误差类型	A	B	C	D	E
单重分形法	最大误差/%	8.41	7.48	13.94	6.04	4.84
	最小误差/%	0.03	0.21	0.00	1.07	0.09
	平均误差/%	4.13	3.69	5.16	3.14	2.39
多重分形法	最大误差/%	4.97	10.09	8.46	9.70	6.86
	最小误差/%	0.11	0.88	0.66	0.34	0.34
	平均误差/%	2.54	5.13	3.77	4.31	3.49
“ARIMA+ 神经网络”法	最大误差/%	43.95	23.72	30.26	59.84	77.91
	最小误差/%	0.54	0.11	0.19	0.52	1.10
	平均误差/%	16.88	6.23	9.80	12.65	24.46

预测方法	误差类型	A	B	C	D	E
单重分形法	最大误差/%	8.41	7.48	13.94	6.04	4.84
	最小误差/%	0.03	0.21	0.00	1.07	0.09
	平均误差/%	4.13	3.69	5.16	3.14	2.39
多重分形法	最大误差/%	4.97	10.09	8.46	9.70	6.86
	最小误差/%	0.11	0.88	0.66	0.34	0.34
	平均误差/%	2.54	5.13	3.77	4.31	3.49
“ARIMA+ 神经网络”法	最大误差/%	43.95	23.72	30.26	59.84	77.91
	最小误差/%	0.54	0.11	0.19	0.52	1.10
	平均误差/%	16.88	6.23	9.80	12.65	24.46

地区	平均误差/%	h
E	2.39	0.9970
D	3.14	0.9805
B	3.69	0.9846
A	4.13	0.9408
C	5.16	0.8961