安徽省空气质量现状分析及预测研究——基于VMD-CNN-LSTM混合模型

An Analysis of the Current Situation and a Predictive Study on Air Quality in Anhui Province—Based on VMD-CNN-LSTM Hybrid Model

摘要: 在全球城市化与工业化快速发展的进程中，空气质量已成为衡量城市可持续发展和居民生活品质的关键指标。本研究聚焦安徽省空气质量，研究发现时间上季度变化呈“U”型，空间上各地市污染物平均浓度差异大。同时通过Pearson相关性分析可知，空气中主要污染物与AQI指数紧密相关。最后构建VMD-CNN-LSTM混合模型预测空气质量，结果显示由于空气质量影响因素复杂、预测时间跨度大带来的不确定因素导致预测存在误差。本研究为安徽及其他区域空气质量管控提供了数据支持与决策参考。

Abstract: In the process of rapid global urbanization and industrialization, air quality has become a key indicator for measuring urban sustainable development and residents’ quality of life. This study focuses on the air quality in Anhui Province and finds that the quarterly variation shows a “U” shape in terms of time, while there are significant differences in the average concentrations of pollutants among various prefecture-level cities in terms of space. Meanwhile, through Pearson correlation analysis, it is found that the main air pollutants are closely related to the AQI index. Finally, a VMD-CNN-LSTM hybrid model is constructed to predict air quality. The results show that there are errors in the prediction due to complex influencing factors of air quality and uncertain factors caused by the large time span of prediction. This study provides data support and decision-making references for air quality management and control in Anhui and other regions.