International Conference on Sensing, Measurement & Data Analytics in the era of Artificial Intelligence (ICSMD2023)

4 / 2023-08-03 21:15:45

Carbon price forecasting method based on CEEMDAN-SE-LiESN

CEEMDAN,SE,LiESN,Carbon price forecasting

Special Sessions > SS10- Artificial Intelligence in Environmental Science

To improve the prediction accuracy of the carbon price, this paper proposes a combined prediction model (CEEMDAN-SE-LiESN) based on the complete integrated empirical modal decomposition (CEEMDAN) with adaptive white noise, sample entropy (SE), and leaky-integrator echo state networks (LiESN). First, the carbon price data are decomposed using CEEMDAN; then, the modal components obtained from the decomposition are reconstructed using sample entropy to get three different sub-sequences; finally, the three sub-sequences are predicted separately using LiESN, and the final prediction results are obtained by integration. The experiments are conducted using the price data of the Chongqing carbon market, and the results show that the proposed model has high prediction accuracy.