为探究皖北城市群大气污染物在不同地域、不同时间下的分布规律以及影响空气中PM_2.5浓度的相关变量，结合2018—2021年安徽省生态环境厅统计数据及相关地域资料，采用主成分分析（principal component analysis, PCA）法按时间段长短及季节性变化分别选取月度数据与年度数据对空气质量的影响因子做相关性分析，对比分析不同季节下空气污染物PM_2.5、PM₁₀的浓度及其他空气污染物浓度的变化，构建基于PCA算法的反向传播神经网络（back-propagation, BP），建立PCA-BP模型并采用交叉-验证法提高模型精度，对大气污染物PM_2.5的浓度做短期预测。实验结果表明：PM_2.5浓度的主要影响因子为PM₁₀、CO、NO₂、SO₂；皖北地区PM_2.5含量整体在冬季偏高；预测模型的精度在夏季与秋季较高，冬季较低，四季的预测精度R²分别达到0.924、0.958、0.935、0.794。

In order to explore the distribution laws of air pollutants in different regions and at different times in the northern Anhui urban agglomeration and the relevant variables affecting the concentration of PM_2.5 in the air, combined with the statistical data of Anhui Provincial Department of Ecology and Environment from 2018 to 2021 and relevant regional data, the principal component analysis (PCA) method was used to analyze the correlation on the influencing factors of air quality by selecting monthly data and annual data respectively according to the time period and seasonal changes. The changes in the concentration of air pollutants PM_2.5 and PM₁₀ and other air pollutants in different seasons were compared and analyzed, and the backpropagation neural network (BP) based on PCA algorithm was constructed. A PCA-BP model was established and a cross-validation method was used to improve the accuracy of the model, and short-term prediction of PM_2.5 concentration in the atmosphere is made. The experimental results show that the main influencing factors of PM_2.5 concentration are PM₁₀, CO, NO₂ and SO₂. PM_2.5 content in northern Anhui is higher in winter on the whole. The accuracy of the prediction model is higher in summer and autumn, but lower in winter. The prediction accuracy R2 of the four seasons reaches 0.924, 0.958, 0.935 and 0.794, respectively.