Method for Predicting Carbon Prices in Civil Aviation Based on Volatility and Trend
"Double carbon"target has become an important and universal value in the international community.Market measures are an indispensable means.The carbon price has become an important factor for the civil aviation to implement the carbon reduction responsibility and obtain the cost-benefit.Accurate prediction of carbon prices has become a major fundamental issue.This paper proposes a carbon price prediction model based on two-way short-term memory network based on multi-band characteristic engineering,aiming to solve the problem of significant fluctuations of carbon price affected by external uncertainties,and help facilitate the implementation of the"two-car-bon"action of civil aviation.Considering the complex nonlinear time series of carbon price data,the white noise com-plete integrated empirical mode decomposition(complete ensemble empirical mode decomposition with adaptive noise,CEEMDAN)method is used to obtain the data characteristics.Then,the Gaussian white noise(White Gaussian noise,WNG)is added to enter the sample entropy simple sequence component to describe the strong volatility of the original sequence,and then to obtain the final feature engineering.Finally,the carbon price is predicted by the gating cycle unit(Gate Recurrent Unit,GRU).The empirical analysis results show the effectiveness of the prediction model,which can provide some method support for airlines to fulfill the contract.
"Double carbon"targetTime series predictionData decompositionGated recurrent unit
NETL
NSTL
万方数据
