Carbon Emission Prediction Under Regression Model

Yashi Nie

《World Scientific Research Journal》 6卷6期 (2020/06) Pp. 181-191

https://doi.org/10.6911/WSRJ.202006_6(6).0024

摘要

Since the first industrial revolution, the rapid increase of greenhouse gas emissions, such as carbon dioxide, has been considered as the main cause of global climate deterioration. CVS survey shows that the global average concentration of greenhouse gases is still growing, China has surpassed the United States to become the largest carbon emission country, the emission reduction warning has been sounded, low-carbon emission reduction is urgent. Based on the actual situation of our country, this paper analyzes the main factors affecting carbon emissions, and makes suggestions for the low-carbon emission reduction in China. First of all, combined with China's economic development, we analyze the changing trend and reasons of China's energy consumption structure from the perspective of industrial structure. Then, we will establish two main models, namely growth rate regression model and prediction model. In the regression model, we will focus on the analysis and prediction of carbon emissions. In order to simplify the calculation, we use the principal component analysis method to calculate the growth rate of each component according to the existing data and the data of the first 15 years (2000- 2014) in the China Statistical Yearbook. Here, we use the growth rate as the dependent variable, by calculating the contribution rate of each influencing factor, we better select three factors that contribute more than 85% of the cumulative annual growth rate of carbon emissions from the eight influencing factors, namely, international trade, industrial structure and energy consumption structure, and use these three main components to replace the original eight influencing factors for subsequent analysis. Then, we use multiple regression method to calculate the partial regression coefficient relative to each factor, and finally get a better growth rate regression model. In the prediction model, we first establish the prediction model for the three principal components. Using the data from 2000 to 2014, we get three principal component trend equations with high fitting degree by Fourier function fitting, and combine these three equations with the regression model in the previous paper to get our final carbon emission prediction model. In order to verify the reliability of the prediction model, we use the prediction model to simulate the carbon emissions in 2015 and 2016, and compare with the real value. Then, we analyze the error, advantages and disadvantages of the model. Finally, we analyze and summarize the results of the model. Combined with China's current economic development and energy consumption structure, we put forward a resource allocation method based on quantitative analysis and give our low-carbon emission reduction suggestions by discussing the change trend of carbon emissions.

關鍵字

Principal component analysis ； multiple regression ； Fourier function

參考文獻

Wang Bing, Zhang Xian, carbon emission prediction research based on improved principal component regression model, global sustainable development report, gsdr brief No. 37cn, 2015.

Jiang Jinhe, carbon emission measurement and analysis of influencing factors in China, resource science, 1007-7588 (2011) 04-0597-082011.

Fu Zhihong, analysis of calculation methods of implicit problems based on input-output table, research on quantitative economy, technology and economy, f233; f2052018.