碳排放权交易市场是实现碳达峰与碳中和目标的核心政策工具,为充分提取其价格序列 非线性, 非平稳性等复杂特征,文章构建一种基于二次分解聚合策略的碳交易价格预测模型CEEMD-EWT-SM-ELM.首先通过互补集合经验模态分解(CEEMD)将原始碳价格序列分解为不同频率的本征模态函数(IMF);然后利用经验小波变换(EWT)针对IMF1分量进行二次分解;再引入样本熵和最大信息系数(SE-MIC,SM)对所有分解得到的分量进行复杂度和相 关度分析,将其划分为高复杂度分量集和低复杂度分量集;进一步将低复杂度分量集聚合为一个分量,并利用极限学习机(ELM)模型对每条分量进行多步预测;最后将所有分量预测结果线性集成.文章选取广东,湖北和天津三个碳交易市场的价格数据进行实例 验证,实证结果表明:在提前1天, 3天和6天的预测中,文章提出组合模型预测精度明显优于基准模型;同时也证实了二次分解聚合策略能够有效提高碳交易价格的预测效果.
Abstract
The carbon emission trading market is the core policy tool to achieve the goal of carbon peak and carbon neutrality. In order to fully extract the nonlinear and non-stationary characteristics of the price series, our paper constructs a carbon trading price prediction model, CEEMD-EWT-SM-ELM, proposed based on the secondary decomposition and aggregation strategy. Firstly, the original carbon price sequence is decomposed into intrinsic mode functions (IMF) of different frequencies by complementary set empirical mode decomposition (CEEMD). Then, the empirical wavelet transform (EWT) is used to secondarily decompose the IMF1 components. After that, sample entropy and maximum information coefficient (SE-MIC, SM) were introduced to analyze the complexity and relevance of all decomposed components, which were divided into the high complexity component set and the low complexity component set. Furthermore, the low complexity components were aggregated into one component, and the ELM model was used to predict each component. Finally, the prediction results of the reconstruction components were linearly integrated. This paper selects the price data of three carbon trading markets in Guangdong, Hubei, and Tianjin for example verification. The empirical results show that: In the prediction of one, three, and six days in advance, the prediction accuracy of the combined model is better than that of the benchmark model, which in turn proves that the secondary decomposition and aggregation strategy can effectively improve the prediction effect of carbon trading price.
关键词
二次分解 /
样本熵 /
最大信息系数 /
碳价格预测
{{custom_keyword}} /
Key words
Secondary decomposition /
sample entropy /
maximum information coefficient /
carbon price forecast
{{custom_keyword}} /
{{custom_sec.title}}
{{custom_sec.title}}
{{custom_sec.content}}
参考文献
[1] Shi H T, Chai J, Lu Q Y, et al. The impact of China's low-carbon transition on economy, society and energy in 2030 based on CO2 emissions drivers. Energy (Oxford), 2022, 239(PD): 122336.
[2] Benz E, Trück S. Modeling the price dynamics of CO2 emission allowances. Energy Economics, 2009, 31(1): 4-15.
[3] Chevallier J. Nonparametric modeling of carbon prices. Energy Economics, 2011, 33(6): 1267-1282.
[4] Suk J B, Cho H J. Forecasting carbon futures volatility using GARCH models with energy volatilities. Energy Economics, 2013, 40(Nov.): 207-221.
[5] Sheng C, Wang G, Geng Y, et al. The correlation analysis of futures pricing mechanism in China's carbon financial market. Sustainability (Basel, Switzerland), 2020, 12(18): 7317.
[6] 朱帮助, 魏一鸣. 基于GMDH-PSO-LSSVM的国际碳市场价格预测. 系统工 程理论与实践, 2011, 31(12): 2264-2271. (Zhu B Z, Wei Y M. Carbon price prediction based on integration of GMDH, particle swarm optimization and least squares support vector machines. Systems Engineering — Theory & Practice, 2011, 31(12): 2264-2271.)
[7] Atsalakis G S. Using computational intelligence to forecast carbon prices. Applied Soft Computing, 2016, 43: 107-116.
[8] Han M, Ding L, Zhao X, et al. Forecasting carbon prices in the Shenzhen market, China: The role of mixed-frequency factors. Energy, 2019, 171(Mar.): 69-76.
[9] Lü J, Cao M, Wu K, et al. Price volatility in the carbon market in China. Journal of Cleaner Production, 2020, 255(C): 120171.
[10] Wang S Y, Lean Y U, Lai K K. Crude oil price forecasting with TEI@I methodology. Journal of Systems Science and Complexity, 2005, 18(2): 145-166.
[11] 崔焕影, 窦祥胜. 基于EMD-GA-BP与EMD-PSO-LSSVM的中国碳市场价格预测. 运筹与管理, 2018, 27(7): 133-143. (Cui H Y, Dou X S. Carbon market price prediction based on EMD-GA-BP and EMD-PSO-LSSVM. Operations Research and Management Science, 2018, 27(7): 133-143.)
[12] 刘金培, 郭艺, 陈华友, 等. 基于非结构数据流行学习的碳价格多尺 度组合预测. 控制与决策, 2019, 34(2): 279-286. (Liu J P, Guo Y, Chen H Y, et al. Multi-scale combined forecast of carbon price based on manifold learning of unstructured data. Control and Decision, 2019, 34(2): 279-286.)
[13] 张大斌, 蔡超敏, 凌立文, 等. 基于CEEMD与GA-SVR的猪肉价格 集成预测模型. 系统科学与数学, 2020, 40(6): 1061-1073. (Zhang D B, Cai C M, Ling L W, et al. Pork price ensemblle prediction model based on CEEMD and GA-SVR. Journal of Systems Science and Mathematical Sciences, 2020, 40(6): 1061-1073.)
[14] Zhang J, Li D, Hao Y, et al. A hybrid model using signal processing technology, econometric models and neural network for carbon spot price forecasting. Journal of Cleaner Production, 2018, 204(PT.1-1178): 958-964.
[15] Liu Z G, Sun W L, Zeng J J. A new short-term load forecasting method of power system based on EEMD and SS-PSO. Neural Computing and Applications, 2014, 24(3-4): 973-983.
[16] Yin H, Dong Z, Chen Y, et al. An effective secondary decomposition approach for wind power forecasting using extreme learning machine trained by crisscross optimization. Energy Conversion and Management, 2017, 150: 108-121.
[17] Sun W, Huang C. A carbon price prediction model based on secondary decomposition algorithm and optimized back propagation neural network. Journal of Cleaner Production, 2020, 243(C): 118671.
[18] Liu H, Mi X, Li Y. An experimental investigation of three new hybrid wind speed forecasting models using multi-decomposing strategy and ELM algorithm. Renewable Energy, 2018, 123(Aug.): 694-705.
[19] 罗宏远, 王德运, 刘艳玲, 等. 基于二层分解技术和改进极限学 习机模型的PM2.5浓度预测研究. 系统工程理论与实践, 2018, 38(5): 1321-1330. (Luo H Y, Wang D Y, Liu Y L, et al. PM2.5 concentration forecasting based on two-layer decomposition technique and improved extreme learning machine. Systems Engineering — Theory & Practice, 2018, 38(5): 1321-1330.)
[20] 梁小珍, 邬志坤, 杨明歌, 等. 基于二次分解策略和模糊时间 序列模型的航空客运需求预测研究. 中国管理科学, 2020, 28(12): 108-117. (Liang X Z, Wu Z K, Yang M G, et al. Research on air passenger demand forecasting based on secondary decomposition strategy and fuzzy time series model. Chinese Journal of Management Science, 2020, 28(12): 108-117.)
[21] Richman J S, Randall M J. Physiological time-series analysis using approximate entropy and sample entropy. American Journal of Physiology-Heart and Circulatory Physiology, 2000, 278(6): H2039.
[22] Reshef D N, Reshef Y A, Finucane H K, et al. Detecting novel associations in large data sets. Science, 2011, 334(6062).
[23] Shi G, Qin C, Tao J, et al. A VMD-EWT-LSTM-based multi-step prediction approach for shield tunneling machine cutterhead torque. Knowledge-Based Systems, 2021, 228(2): 107213.
[24] Gilles J. Empirical Wavelet Transform. IEEE Transactions on Signal Processing, 2013, 61(16): 3999-4010.
[25] Huang G B, Zhu Q Y, Siew C K. Extreme learning machine: Theory and applications. Neurocomputing, 2006, 70(1-3): 489-501.
[26] Zhang F, Fang H, Song W Y. Carbon market maturity analysis with an integrated multi-criteria decision making method: A case study of EU and China. Journal of Cleaner Production, 2019, 241(C): 118296.
[27] Lu H F, Ma X, Huang K, et al. Carbon trading volume and price forecasting in China using multiple machine learning models. Journal of Cleaner Production, 2020, 249: 119386.
[28] 周坤, 高晓辉, 李廉水.基于EMD-XGB-ELM和FSGM双元处理的碳排放 交易价格集成预测. 中国管理科学, 2022, DOI: 10.16381/j.cnki.issn1003-207x. (Zhou K, Gao X H, Li L S. Integrated carbon emission trading price prediction based on EMD-XGB-ELM and FSGM from the perspective of dual processing. Chinese Journal of Management Science, 2022, DOI: 10.16381/j.cnki.issn1003-207x.
{{custom_fnGroup.title_cn}}
脚注
{{custom_fn.content}}
基金
国家自然科学基金(71971089, 72001083), 广东省基础与应用基础研究基金(2022A1515011612)资助课题.
{{custom_fund}}
PDF(895 KB)
