A Novel Combined Prediction Scheme Based on CNN and LSTM for Urban PM2.5 Concentration

Urban air pollutant concentration prediction is dealing with a surge of massive environmental monitoring data and complex changes in air pollutants. This requires effective prediction methods to improve prediction accuracy and to prevent serious pollution incidents, thereby enhancing environmental management decision-making capacity. In this paper, a new pollutant concentration prediction method is proposed based on the vast amounts of environmental data and deep learning techniques. The proposed method integrates big data by using two kinds of deep networks. This method is based on the design that uses a convolutional neural network as the base layer, automatically extracting features of input data. A long short-term memory network is used for the output layer to consider the time dependence of pollutants. Our model consists of these two deep networks. With performance optimization, the model can predict future particulate matter (PM2.5) concentrations as a time series. Finally, the prediction results are compared with the results of numerical models. The applicability and advantages of the model are also analyzed. The experimental results show that it improves prediction performance compared with classic models.

[1]  Samuel Soret,et al.  The Association between Ambient Fine Particulate Air Pollution and Lung Cancer Incidence: Results from the AHSMOG-2 Study , 2016, Environmental health perspectives.

[2]  Jian Sun,et al.  Spatial Pyramid Pooling in Deep Convolutional Networks for Visual Recognition , 2014, IEEE Transactions on Pattern Analysis and Machine Intelligence.

[3]  Silvio Savarese,et al.  Social LSTM: Human Trajectory Prediction in Crowded Spaces , 2016, 2016 IEEE Conference on Computer Vision and Pattern Recognition (CVPR).

[4]  Giorgio Corani,et al.  Air pollution prediction via multi-label classification , 2016, Environ. Model. Softw..

[5]  Geoffrey E. Hinton,et al.  Learning representations by back-propagating errors , 1986, Nature.

[6]  H. Zou,et al.  Regularization and variable selection via the elastic net , 2005 .

[7]  James M. Wilczak,et al.  PM 2.5 analog forecast and Kalman filter post-processing for the Community Multiscale Air Quality (CMAQ) model , 2015 .

[8]  Geoffrey E. Hinton,et al.  Learning representations by back-propagation errors, nature , 1986 .

[9]  James A Mulholland,et al.  Bayesian-based ensemble source apportionment of PM2.5. , 2013, Environmental science & technology.

[10]  Koji Zettsu,et al.  Dynamically pre-trained deep recurrent neural networks using environmental monitoring data for predicting PM2.5 , 2015, Neural Computing and Applications.

[11]  Wang Jian.  Study on Forecasting of Air Pollution Based on BP Model , 2002 .

[12]  Fei-Yue Wang,et al.  Traffic Flow Prediction With Big Data: A Deep Learning Approach , 2015, IEEE Transactions on Intelligent Transportation Systems.

[13]  Hermann Ney,et al.  From Feedforward to Recurrent LSTM Neural Networks for Language Modeling , 2015, IEEE/ACM Transactions on Audio, Speech, and Language Processing.

[14]  Matthew E. Kahn,et al.  Self-Protection Investment Exacerbates Air Pollution Exposure Inequality in Urban China , 2015 .

[15]  Razvan Pascanu,et al.  Advances in optimizing recurrent networks , 2012, 2013 IEEE International Conference on Acoustics, Speech and Signal Processing.

[16]  A. Szpiro,et al.  Impact of preferential sampling on exposure prediction and health effect inference in the context of air pollution epidemiology , 2015, Environmetrics.

[17]  Erkki Oja,et al.  Time series prediction competition: The CATS benchmark , 2007, Neurocomputing.

[18]  Kaiming He,et al.  Faster R-CNN: Towards Real-Time Object Detection with Region Proposal Networks , 2015, IEEE Transactions on Pattern Analysis and Machine Intelligence.

[19]  Jürgen Schmidhuber,et al.  Deep learning in neural networks: An overview , 2014, Neural Networks.

[20]  J. DiCarlo,et al.  Using goal-driven deep learning models to understand sensory cortex , 2016, Nature Neuroscience.

[21]  Rina Dechter,et al.  Learning While Searching in Constraint-Satisfaction-Problems , 1986, AAAI.

[22]  Yoon Kim,et al.  Convolutional Neural Networks for Sentence Classification , 2014, EMNLP.

[23]  Jürgen Schmidhuber,et al.  Long Short-Term Memory , 1997, Neural Computation.

[24]  Yuan Zhang,et al.  Short-Term Residential Load Forecasting Based on LSTM Recurrent Neural Network , 2019, IEEE Transactions on Smart Grid.

[25]  Dong Yu,et al.  Deep Learning: Methods and Applications , 2014, Found. Trends Signal Process..

[26]  Soon-Bark Kwon,et al.  Predicting PM10 concentration in Seoul metropolitan subway stations using artificial neural network (ANN). , 2018, Journal of hazardous materials.

[27]  Kunikazu Kobayashi,et al.  Time series forecasting using a deep belief network with restricted Boltzmann machines , 2014, Neurocomputing.

[28]  Yoshua Bengio,et al.  Gradient-based learning applied to document recognition , 1998, Proc. IEEE.

[29]  Qi Li,et al.  A Spatiotemporal Prediction Framework for Air Pollution Based on Deep RNN , 2017 .

[30]  M. Brauer,et al.  Transboundary health impacts of transported global air pollution and international trade , 2017, Nature.