Highway Safety Warning System in Heavy Air Pollution Environment: An Example from Xi’an, China

Air pollution is one of the hot and challenging issues in urban management, and it significantly impacts human health and the functioning of urban infrastructure. Air quality prediction can be an effective tool to obtain air quality in advance after a particular time in the future. However, existing methods fail to adequately consider the impact of spatial correlation characteristics of air quality monitoring stations on prediction accuracy. This study proposes an extended- and short-term memory prediction model considering the spatial correlation of monitoring stations to analyze the spatial correlation and predict pollutant concentrations at six air quality monitoring stations in Xi’an city. The results show that the new model has better stability and performance than the old models and the correlation coefficient R² (0.70−0.92). The model is applied to air pollutant concentration prediction to provide data support for highway safety warning systems operation in a heavy pollution environment.