The Construction And Analysis Of The Chinese Carbon Price Prediction Model Based On EEMD-BILSTM

Since human stepped into industrial society,the global climate environment has been deteriorating day by day.Since the 21 st century,global warming has been particularly severe.Carbon emission has become a great challenge faced by all countries in the world.In the report of the 20 th National Congress of the Communist Party of China,it was once again emphasized that China’s dual carbon goals should be actively and steadily promoted to reach carbon peak and carbon neutrality,and specific arrangements should be made for improving the carbon market.In order to rationally plan carbon emission and establish carbon emission order,the EU gave birth to the world’s first carbon emission trading market in 2005.In 2013,starting from Shenzhen,China has successively opened 9 Pilotcarbon trading markets.Prediction of carbon price is an essential part of carbon trading activities.Based on the research of traditional carbon price prediction models,a new carbon price prediction model is developed to achieve accurate prediction of carbon price with the aim of improving the robustness,accuracy and universal applicability of carbon price prediction models.As for the theoretical aspects of the model,This article mainly adopts the method of combining signal decomposition algorithm with deep learning algorithm,And innovatively applies the Ensemble Empirical Mode Analysis Algorithm and the Bidirectional Long and Short Term Memory Neural Networks to the Field of Carbon Trading Price Prediction,which is a new attempt of this method in this field.This document is mainly divided into the following sections: First of all,this article introduces the background and meaning of the study,domestic and international research status and literature review;Then it introduces the decomposition algorithm,deep learning algorithm principle and related theory of improved algorithm,data acquisition and preprocessing.The last part is the empirical analysis.Firstly,In this paper,the EEMD sets are decomposed into several parts by the experience mode decomposition,Then,the reconstructed parts are placed in the bidirectional long and short term memory neural network model for the purpose of fitting and forecasting.Finally,the proposed model is compared with the traditional model,which proves its robustness,precision and universality.In conclusion,this paper draws the following conclusions: First,it is feasible to apply the EEMD-Bi LSTM model in the field of carbon trading price prediction,which has reference significance for improving the accuracy of carbon price prediction.Second,EEMD-Bi LSTM model performs well in robustness,accuracy and general applicability,Which is of relevance to improving the accuracy of carbon price forecasting.Thirdly,according to the forecast results of the model,the paper gives some advice for developing the trading market.