【作者】 莫娜（Mona Ahmed Omer）；

【导师】 张莹；

【作者基本信息】 华北电力大学(北京) ， 计算机应用技术， 2020， 硕士

【摘要】 近年来,人们越来越关注空气质量,因为它直接影响人们的健康和日常生活。有效的空气质量预测已成为研究的热点之一。在现代社会中,空气污染是一个重要的话题,因为这种污染严重影响人类健康和环境。在空气污染物中,颗粒物(PM2.5)由直径等于或小于2.5 μm的悬浮颗粒组成。PM2.5的来源可能是燃煤发电,烟雾或粉尘。空气中的这些悬浮颗粒会损害人体的呼吸系统和心血管系统,从而可能进一步导致其他疾病,例如哮喘,肺癌或心血管疾病。本文面临着许多挑战,例如数据源的不稳定性以及污染物浓度随时间序列的变化。针对这一问题,我们提出了一种基于深度学习的改进的空气质量预测方法,来预测下一个小时北京35个空气质量监测站的PM2.5浓度。本文我们通过创建一个能够预测PM2.5浓度的新框架来解决此问题,我们将创建一个深度学习模型,以监视和估计PM2.5浓度,将卷积神经网络(CNN)和长期短期记忆(LSTM)结合,并将其应用于PM2.5浓度的预测。为了比较每种算法的整体性能,本文在实验中采用了平均绝对误差(MAE)、均方根误差(RMSE)、对称平均绝对百分比误差(SMAPE)三种测量指标。实验结果表明,与其他机器学习方法相比,本文提出的CNN-LSTM模型的预测精度最高。对于CNN-LSTM模型,本文还验证了其预测PM2.5浓度的可行性和实用。将来,这项研究也可以应用于预防和控制PM2.5。更多还原

【Abstract】 In recent years,people have been paying more and more attention to air quality because it directly affects people’s health and daily life.Effective air quality prediction has become one of the hot research issues.In modern society,air pollution is an important topic as this pollution exerts a critically bad influence on human health and the environment.Among air pollutants,Particulate Matter(PM2.5)consists of suspended particles with a diameter equal to or less than 2.5 μm.Sources of PM2.5 can be coal-fired power generation,smoke,or dusts.These suspended particles in the air can damage the respiratory and cardiovascular systems of the human body,which may further lead to other diseases such as asthma,lung cancer,or cardiovascular diseases.This thesis is facing many challenges,such as the instability of data sources and the variation of pollutant concentration along time series.Aiming at this problem,we propose an improved air quality prediction method based on the deep learning to predict the PM2.5 concentration at the 35 air quality monitoring stations in Beijing over the next hour.In this thesis we will address this problem by creating a new framework capable of predicting the PM2.5 concentration,for this purpose we will create a deep learning model,to monitor and estimate the PM2.5 concentration,Convolutional Neural Network(CNN)and Long Short-Term Memory(LSTM)are combined and applied to forecast PM2.5 concentration.To compare the overall performance of each algorithm,three measurement indexes,Mean Absolute Error(MAE),Root Mean Square Error(RMSE),Symmetric Mean Absolute Percentage Error(SMAPE)are applied to the experiments in this thesis.Compared with other machine learning methods,the experimental results showed that the forecasting accuracy of the proposed CNN-LSTM model is verified to be the highest in this thesis.For the CNN-LSTM model,its feasibility and practicability to forecast the PM2.5 concentration are also verified in this thesis.In the future,this study can also be applied to the prevention and control of PM2.5.更多还原