基于灰色-BP神经网络安徽省二氧化碳排放量预测研究
Prediction of Carbon Dioxide Emission in Anhui Province Based on Grey-BP Neural Network

为顺利实现2030年前碳达峰目标，研究安徽省二氧化碳排放量的主要影响因素和预测模型可促进其制定合理的减碳政策，本文先使用灰色关联系数用于筛选出安徽省二氧化碳排放主要影响因素，构建BP神经网络中的输入层，再利用训练完成的BP神经网络预测安徽省二氧化碳量。选取近23年数据进行相关验算，预测结果表明：1) 与GM (1, 1)和BP神经网络模型进行对比，组合预测模型平均绝对百分比误差为1.14%，该模型具有良好的预测效果；2) 利用该模型预测出安徽省2023至2027年二氧化碳排放量，与往年相比，二氧化碳增加较少，趋于稳定，得到一定的控制。
In order to achieve the carbon peak goal before 2030, the study on the main influencing factors and prediction models of carbon dioxide emissions in Anhui Province can promote the development of reasonable carbon reduction policies. In this paper, the grey correlation coefficient is first used to screen out the main influencing factors of carbon dioxide emissions in Anhui Province, and the input layer of BP neural network is constructed. The trained BP neural network was used to predict the carbon dioxide quantity in Anhui Province. The results show that: 1) Compared with GM (1, 1) and BP neural network models, the average absolute percentage error of the combined prediction model is 1.14%, and the model has a good prediction effect; 2) Using this model, it is predicted that carbon dioxide emissions in Anhui Province from 2023 to 2027 will increase less than in previous years, and tend to be stable and under certain control.