%0 Journal Article
%T 基于iTransformer模型的碳排放权价格预测——以广东碳排放权为例
Carbon Emission Rights Price Prediction Based on the iTransformer Model—A Case Study of Guangdong Carbon Emission Rights
%A 张明赫
%A 袁中华
%J Journal of Low Carbon Economy
%P 127-136
%@ 2324-7932
%D 2025
%I Hans Publishing
%R 10.12677/jlce.2025.141014
%X 在世界主要经济体高度关注生态环境问题，我国坚持绿色发展战略，努力实现“双碳”目标的背景下，碳排放权市场成为了一项有效的减排政策工具，被多国所使用。本研究聚焦碳排放权的交易价格，从多角度分析价格影响因素，构建iTransformer机器学习模型来分析价格波动规律及走势，以广东省碳排放GDEA为实证案例进行价格预测，并使用SHAP分析展示了各因素的影响方式和重要性。通过与主流的LSTM模型预测结果进行对比分析，预验证了iTransformer模型的泛用性及预测效果的准确性和可靠性。根据实证研究结果，本文为未来研究方向提供了一些思路，为我国碳市场发展提供了三点政策建议。
With the global focus on ecological and environmental issues among major economies, China adheres to a green development strategy and strives to achieve its “dual carbon” goals. Under this context, the carbon emission rights market has become an effective policy tool for emission reduction, adopted by many countries. This study concentrates on the trading price of carbon emissions, analyzing the influencing factors from multiple perspectives. It constructs an iTransformer machine learning model to examine price fluctuation patterns and trends, using the carbon emissions GDEA of Guangdong Province as an empirical case for price prediction. Furthermore, SHAP analysis is employed to illustrate the impact and significance of various factors. By comparing the prediction results with those of the mainstream LSTM model, this study preliminarily verifies the versatility and accuracy of the iTransformer model. Based on empirical findings, the paper offers insights into future research directions and provides three policy recommendations for the development of China’s carbon market.
%K 广东碳排放权，
%K 时间序列分析
Guangdong Carbon Emission Rights
%K Time Series Analysis
%K iTransformer
%K LSTM
%U http://www.hanspub.org/journal/PaperInformation.aspx?PaperID=108069