Air Quality Prediction Model Using Deep Learning in Internet of Things Environmental Monitoring System

In order to realize the accurate prediction of spatial-temporal air quality index, this paper constructs a STAQI prediction model based on deep learning, including data processing, spatial feature acquisition, temporal feature acquisition, and STAQI prediction. Firstly, the spatial interpolation method is used to optimize the sample data set to provide reliable data; the improved graph convolutional network and the improved long short-term memory are used to effectively extract the spatial and temporal distribution characteristics of AQI data; and then, the extreme learning machine model is used to accurately predict and analyze AQI data. Simulation results show that the evaluation indexes RMSE and MAE of the constructed prediction model are 4.51 and 3.92, respectively, showing excellent curve fitting ability and AQI prediction ability.