摘要
碳交易价格受到宏观经济、能源政策等多种因素的影响,表现出强波动性、非线性等特征,给碳交易价格的准确预测带来巨大困难。针对这一问题,基于二次分解和误差修正策略构建一种碳交易价格预测模型:首先,使用浣熊优化算法优化的变分模态分解方法分解碳价序列,降低原始序列的复杂度;其次,使用经验小波变换对变分模态分解产生的残差序列进行二次分解,充分提取残差序列中的有效信息;然后,使用浣熊优化算法优化的极限学习机对各分量进行预测,获得初始预测结果和误差序列;最后,使用基本和浣熊优化算法优化的极限学习机对误差序列进行分解和预测,并利用误差预测结果对初始预测结果进行修正,得到最终预测结果。选取深圳、湖北和福建3个碳交易市场的碳价数据进行实证验证,结果表明,所提出的模型相比于其他对照模型具有更优异的预测精度和稳定性,有效提高碳价预测的准确性。
The carbon trading price is affected by many factors such as macroeconomic and energy policies,showing strong volatility and nonlinearity,which poses great difficulties for accurate forecasting of carbon trading prices.To address this issue,a carbon trading price forecasting model is constructed based on the secondary decomposition and error correction strategy.Firstly,the carbon price sequence is decomposed by the variational mode decomposition(VMD)optimized by the coati optimization algorithm(COA)to reduce the sequence complexity.Secondly,the empirical wavelet transform(EWT)is used to decompose the residual sequence generated by VMD to fully extract the effective information in the sequence.Then,the extreme learning machine(ELM)optimized by COA is used to forecast each component,and the initial forecasting results and error sequence are obtained.Finally,EWT and COAELM are used to decompose and forecast the error sequence,and the initial forecasting results are corrected by the error forecasting results to obtain the final forecasting results.The carbon price data of three carbon trading markets in Shenzhen,Hubei and Fujian are selected for empirical verification.The results show that the model has better forecasting accuracy and stability than other control models,and effectively improves the accuracy of carbon price prediction.
作者
何志超
黄建华
He Zhichao;Huang Jianhua(School of Economics and Management,Fuzhou University,Fuzhou 350108,China)
出处
《科技管理研究》
2024年第13期200-214,共15页
Science and Technology Management Research
基金
国家社会科学基金一般项目“面向城乡双向流通的物流网络布局优化及多主体协同服务机制研究”(20BGL003)。
关键词
碳交易价格
二次分解
浣熊优化算法
极限学习机
误差修正
carbon trading price
secondary decomposition
coati optimization algorithm
extreme learning machine
error correction
