深度学习在PM2.5预测中的应用
|
摘要
近年来,随着我国环境保护措施的不断加强,如何有效控制大气污染也越来越重要,而PM2.5作为污染物的重要来源也提上日程,因此对PM2.5做有效的预测具有重要的现实意义和社会价值。收集2015年1月1日到2019年3月10日昆明市空气质量指数信息,首先利用传统的时间序列模型预测PM2.5,接下来设计LSTM循环神经网络PM2.5预测模型。实验结果表明:与传统的时间序列序列相比,LSTM模型预测精度更高,具有较强的实用性。
In recent years,with the continuous strengthening of China's environmental protection measures,how to effectively control air pollution is becoming more and more important,and PM2.5 as an important source of pollutants has also been put on the agenda,so the effective prediction of PM2.5 has Important practical and social values.Collects information on Kunming air quality index from January 1,2015 to March10,2019.First,uses the traditional time series model to predict PM2.5.Next,designs the LSTM cyclic neural network PM2.5 prediction model.The experimental results show that the LSTM model has higher prediction accuracy than the traditional time series,has a strong practicality.
In recent years,with the continuous strengthening of China's environmental protection measures,how to effectively control air pollution is becoming more and more important,and PM2.5 as an important source of pollutants has also been put on the agenda,so the effective prediction of PM2.5 has Important practical and social values.Collects information on Kunming air quality index from January 1,2015 to March10,2019.First,uses the traditional time series model to predict PM2.5.Next,designs the LSTM cyclic neural network PM2.5 prediction model.The experimental results show that the LSTM model has higher prediction accuracy than the traditional time series,has a strong practicality.
引文
[1]Feng X,Li Q,Zhu Y,et al.Artificial Neural Networks Forecasting of PM2.5,Pollution Using Air Mass Trajectory Based Geographic Model and Wavelet Transformation[J].Atmospheric Environment,2015,107:118-128.
[2]王静,杨复沫,王鼎益,等.北京市MODIS气溶胶光学厚度和PM2.5质量浓度的特征以及相关性[J].中国科学院大学学报,2010,27(1):10-16.
[3]孙永霞.基于主成分-BP神经网络模型的雾霾天数预测的研究[D].江苏大学,硕士论文,2015.
[4]颜玉倩,朱克云,张杰.成都地区春季一次持续性灰霾天气过程特征及预测[J].气象与环境学报,2016,32(1):33-39.
[5]Sainath,T.N.,Vinyals,O.,Senior,A.,Sak,H.Convolutional,Long Short-Term Memory,Fully Connected Deep Neural Network.In:2015IEEE international Conference on Acoustics,Speech and Signal Processing(ICASSP).IEEE April 2015:4580-4584.
[6]Mantas Miksys.Predictions of PM2.5 and PM10 Concentrations Using Static and Mobile Sensors,4th Year Project Report Computer Science School of Informatics University of Edinburgh[C],2016.
[7]Vikram Reddy,Pavan Yedavalli,Shrestha Mohanty,Udit Nakhat.Deep Air:Forecasting Air Pollution in Beijng,China[C],2017.
[8]魏贞原.深度学习[M].北京:电子工业出版社,2018.5:177-192.
[9]王燕.时间序列分析-基于R[M].北京:中国人民大学出版社,2015.9:47-69.
[10]Robert I.Kabacoff,著,王小宁,黄俊文,等译.R语言实战[M].第二版.2016.11:333-319.
[11]周永生.基于LSTM模型的PM2.5预测[D].湖南:湖南大学,2018,5.
[12]谢小峰.面向物联网时间序列数据深度学习LSTM学习方法研究[D].湖南:湖南大学,2018,5.
[13]吴茂贵,王冬,李涛,杨本法.Python深度学习[M].北京:机械工业出版社,2018.10:226-234.
[2]王静,杨复沫,王鼎益,等.北京市MODIS气溶胶光学厚度和PM2.5质量浓度的特征以及相关性[J].中国科学院大学学报,2010,27(1):10-16.
[3]孙永霞.基于主成分-BP神经网络模型的雾霾天数预测的研究[D].江苏大学,硕士论文,2015.
[4]颜玉倩,朱克云,张杰.成都地区春季一次持续性灰霾天气过程特征及预测[J].气象与环境学报,2016,32(1):33-39.
[5]Sainath,T.N.,Vinyals,O.,Senior,A.,Sak,H.Convolutional,Long Short-Term Memory,Fully Connected Deep Neural Network.In:2015IEEE international Conference on Acoustics,Speech and Signal Processing(ICASSP).IEEE April 2015:4580-4584.
[6]Mantas Miksys.Predictions of PM2.5 and PM10 Concentrations Using Static and Mobile Sensors,4th Year Project Report Computer Science School of Informatics University of Edinburgh[C],2016.
[7]Vikram Reddy,Pavan Yedavalli,Shrestha Mohanty,Udit Nakhat.Deep Air:Forecasting Air Pollution in Beijng,China[C],2017.
[8]魏贞原.深度学习[M].北京:电子工业出版社,2018.5:177-192.
[9]王燕.时间序列分析-基于R[M].北京:中国人民大学出版社,2015.9:47-69.
[10]Robert I.Kabacoff,著,王小宁,黄俊文,等译.R语言实战[M].第二版.2016.11:333-319.
[11]周永生.基于LSTM模型的PM2.5预测[D].湖南:湖南大学,2018,5.
[12]谢小峰.面向物联网时间序列数据深度学习LSTM学习方法研究[D].湖南:湖南大学,2018,5.
[13]吴茂贵,王冬,李涛,杨本法.Python深度学习[M].北京:机械工业出版社,2018.10:226-234.