基于消光系数的机场PM2.5质量浓度神经网络预测模型

分析了气溶胶粒径吸湿增长因子、风速和NO_2与消光系数和PM_(2.5)质量浓度之间的相关性及影响规律。提出了一种基于消光系数的机场PM_(2.5)质量浓度神经网络预测模型。首先,建立消光系数与PM_(2.5)质量浓度之间的定量关系,并分析相对湿度对其影响。然后,分析风速和NO_2对消光系数和PM_(2.5)质量浓度的影响。最后,将四项参数与PM_(2.5)质量浓度之间的复杂关系通过模糊神经网络进行学习和表达,实现PM_(2.5)质量浓度的预测。使用实测PM_(2.5)质量浓度数据对预测模型进行了对比验证。结果表明,该预测模型的预测精度较高,能较为客观的反映机场PM_(2.5)质量浓度的变化情况,这对研究颗粒物质量浓度对机场能见度的影响规律以及机场周边污染治理决策提供数据支持具有重要的意义。

Neural Network Prediction Model of Particulate Matter Concentration around Airport Based on Extinction Coefficient

In the paper the correlation between Aerosol hygroscopic growth factor, wind speed and air emission and extinction coefficient and PM2.5 mass concentration was analyzed, a neural network prediction model based on the extinction coefficient of the airport PM2.5 mass concentration is proposed. Firstly, the quantitative relationship between the extinction coefficient and the mass concentration of PM2.5 was established. Then, the influence of wind speed and extinction coefficient and PM2.5 mass concentration was analyzed. Finally, the complex relationship between the four parameters and the mass concentration of PM2.5 is studied and expressed by fuzzy neural network to realize the prediction of PM2.5 mass concentration. The prediction model is validated by the measured PM2.5 mass concentration data. The results show that the prediction accuracy of the model is high, and it can objectively reflect the change of PM2.5 mass concentration, It is of great significance to study the influence of the mass concentration of the particulate matter on the visibility of the airport and to provide data support for the decision making of the pollution control around the airport.