基于Transformer-LSTM模型的多因素碳排放权交易价格预测

Multi-factor carbon emission rights trading price prediction based on Transformer-LSTM model

碳排放权交易作为一种重要的环境政策工具在全球范围内得到了广泛应用。如何运用深度学习等技术提高碳排放权价格预测能力是一个重要问题,基于此,提出一种Transformer-LSTM多因素碳排放权交易价格预测的深度学习模型,以湖北省碳排放权交易价格为例,旨在探索运用深度学习的方法,预测湖北省碳排放权交易价格的变动趋势。输入Transformer-LSTM模型进行预测,同时运用支持向量机回归(SVR)、多层感知机(MLP)、长短时记忆网络(LSTM)、Transformer模型进行预测与对比。通过在历史数据上进行训练,实验结果表明,Transformer-LSTM模型得到的预测价格与湖北省碳排放权交易价格(HBEA)的实际价格更为吻合,在平均绝对误差(MAE)、均方误差(MSE)、均方根误差(RMSE)和评估指标上也有更佳的表现。

Carbon emission trading,as an important environmental policy tool,has been widely used in the world.How to use deep learning and other technologies to improve the predictive ability of carbon emission rights prices is an important issue.Based on this,this paper proposes a deep learning model for multi-factor carbon emission rights trading price prediction with Transformer-LSTM.Taking the carbon emission trading price in Hubei Province as an example,this paper aims to explore the use of deep learning methods to predict the changing trend of carbon emission trading price in Hubei Province.We input Transformer-LSTM model for prediction,and use support vector regression(SVR),multilayer perceptron(MLP),long short-term memory(LSTM)and Transformer model for prediction and comparison.Through training on the historical data,the experimental results show that the predicted prices obtained by the Transformer-LSTM model are more consistent with the actual price of Hubei carbon emission allowance(HBEA),and perform better in terms of mean absolute error(MAE),mean square error(MSE),root mean square error(RMSE),and evaluation indicators.