Carbon emission reduction prediction of new energy vehicles in China based on GRA-BiLSTM model
Bingchun Liu,Shuai Wang,Xiaoqin Liang,Zhaoyang Han
DOI: https://doi.org/10.1016/j.apr.2023.101865
IF: 4.831
2023-09-01
Atmospheric Pollution Research
Abstract:In response to the greenhouse effect, 178 countries (regions) around the world have signed “the Paris Agreement” to combat climate change. As the world's second largest source of carbon emissions, the transport sector is in urgent need of “green transformation”. China is working to reduce carbon emissions from transport by developing a new energy vehicle (NEV) industry. In order to ensure the accurate formulation and promotion of government policies, accurate prediction of NEV ownership is crucial. To this end, this study developed a combined model based on grey relational analysis and bi-directional long- and short-term memory (GRA-BiLSTM). Firstly, GRA was used to evaluate and screen the experimental data indicators that affect NEV retention. Secondly, BiLSTM model was used to learn the characteristics of important impact indicators. The mean absolute percentage error (MAPE) of the GRA-BiLSTM combined model established in this study is 5.16%, which is lower than the other seven comparative prediction models. Then, three development scenarios of low, medium, and high are set to predict the new energy vehicle ownership in China from 2020 to 2030 and calculate the carbon emission reduction. The results show that in the three development scenarios of low, medium and high, the new energy vehicle ownership develops to 35,228.08, 51,865.48 and 71,887.82 thousand vehicles in 2030, respectively, and the calculated carbon emission reduction quantities are 3433835.63 Metric Tons, 4600719.93 Metric Tons, and 5837315.76 Metric Tons, respectively. Finally, the NEV retention and carbon emission reductions for 2031–2060 are projected based on the average development scenarios.
environmental sciences