Air Quality Forecasting Modeling MethodBased on CNN-LSTM
With human activities and natural evolution,various substances are emitted into the atmosphere and accumulate to certain concentrations,causing air pollution that harms the ecolog-ical environment and health.Therefore,establishing air quality forecasting models to predict and control air pollution in advance is an effective method to improve air quality.This paper utilizes feature engineering to categorize air quality input data into temporal and spatial features and conducts layered modeling.A CNN-LSTM network structure is then introduced,where spatial variables are input into the CNN,and the spatial variables output by the convolution operations along with the temporal variables are input into the LSTM to deeply explore the spatiotemporal characteristics of the data.Experimental results show that the model performs well,has strong scalability,and is effective in exploring the intrinsic connections within meteorological data.
air qualitypredictive modelingdata mininglong short-term memory networkconvolutional neural network
NETL
NSTL
万方数据
