摘要
对碳价进行科学准确的预测,有利于提高碳市场风险防范能力。本文提出一种基于二次分解和机器学习的碳价组合预测模型,将影响碳交易价格的内部和外部因素输入预测模型,汇总各序列的最优预测值即得到碳价的最终组合预测结果。以湖北碳市场价格为样本开展实证研究,结果表明:该组合预测模型达到了较高的预测精度,有利于提高碳交易价格预判的准确度、提高碳市场运行效率。基于此,建议采取二次分解和机器学习方法开展碳交易价格预测工作,建立跨试点碳价风险预警联动机制,根据经济发展形势及时调整碳配额发放总量。
Accurate carbon price prediction is conducive to improve the carbon market's risk prevention ability.This paper proposed a carbon price combined prediction model based on quadratic decomposition and machine learning.Both the internal and external influenc-ing factors were inputted into the prediction mode,and the final combined prediction result of carbon price can be obtained by aggregating the optimal predicted value of each sequence.Taking Hubei carbon market price as a sample,the empirical result showed the combined prediction model proposed achieved high prediction accuracy,which is beneficial to improve carbon trading price prediction accuracy and carbon market operation efficiency.Based on this,it is suggested to adopt quadratic decomposition and machine learning methods to forecast carbon trading price,establish cross-pilot carbon price risk warning linkage mechanism,and adjust the total amount of carbon allowances ac-cording to the economic development situation timely.
出处
《价格理论与实践》
北大核心
2023年第9期142-145,209,共5页
Price:Theory & Practice
基金
河北省社科基金项目“‘十四五’时期河北省能源消费结构目标优化及政策模拟”(HB20GL027)
中央高校基本科研业务费专项资金资助(2021MS105)。
关键词
“双碳”目标
碳交易
碳价预测
二次分解
机器学习
“dual carbon”target
carbon trading
carbon price prediction
quadratic decomposition
machine learning
